Wednesday 18 March 2015

How do I learn machine learning?



  • Start from Series of videos ::https://www.youtube.com/view_play_list?p=A89DCFA6ADACE599
  •  There were plenty of ways to learn Machine Learning / data mining.
    1) Primary thing is to select and learn most suitable language for data mining. ex: python, R.
    I prefer python.

    2)Learning some packages like numpy,pandas,matplotlib,scikit-learn.

    3) Just follow any reference book for data mining / Machine Learning.

    4) Just learning the algorithm will never make you perfect person in data mining or Machine learning field you have to apply them for some data sets. you can get data sets from internet itself.

    5) If you face any problem while applying your algorithm to data sets. Just go to A Programmer's Guide to Data Mining here you can find some very good example of how to apply data mining algorithms to data sets and with some real world  applications. I promise you if your an  really  datamining / Machine learning  aspirant you will enjoy A Programmer's Guide to Data Mining.

    6) Once you completed this now you can search for data sets and you can apply your algorithms.

    7) You can spent some time on  Coursera. where you will find some coursers belongs to data mining and Machine learning. (you will get videos and problem to practice)  

    8) On a final note i can say that quora + google will give you lots of stuff related to data mining.

    9) All the best ,have a blast and have fun play with data mining / Machine Learning .

    10) You can check this blog once a week.
    dataaspirant
  • For a more gentle read you may want to check out the textbook by T.Mitchell
  • The most coherent and concise lecture notes are by Andrew Ng,   also see his ML course
  • For Natural Language applications see Michael Collins'  NLP course and notes
  • For Graphical Models see Daphne Koller's course and book
  • For Deep Learning see NN + UFLDL + CNN courses

Wednesday 7 January 2015

How do I learn about big data?

I'll try to give a very crude overview of how the pieces fit in together, because the details span multiple books. Please forgive me for some oversimplifications.

  • MapReduce is the Google paper that started it all (Page on googleusercontent.com). Its a paradigm for writing distributed code inspired by some elements of functional programming. You don't have to do things this way, but it neatly fits a lot of problems we try to solve in a distributed way. The Google internal implementation is called MapReduce and Hadoop is it's open-source implementation. Amazon's Hadoop instance is called Elastic MapReduce (EMR) and has plugins for multiple languages.
  • HDFS is an implementation inspired by the Google File System (GFS) to store files across a bunch of machines when it's too big for one. Hadoop consumes data in HDFS (Hadoop Distributed File System).
  • Apache Spark is an emerging platform that has more flexibility than MapReduce but more structure than a basic message passing interface. It relies on the concept of distributed data structures (what it calls RDDs) and operators. See this page for more: The Apache Software Foundation
  • Because Spark is a lower level thing that sits on top of a message passing interface, it has higher level libraries to make it more accessible to data scientists. The Machine Learning library built on top of it is called MLib and there's a distributed graph library called GraphX.
  • Pregel and it's open source twin Giraph is a way to do graph algorithms on billions of nodes and trillions of edges over a cluster of machines. Notably, the MapReduce model is not well suited to graph processing so Hadoop/MapReduce are avoided in this model, but HDFS/GFS is still used as a data store.
  • Zookeeper is a coordination and synchronization service that a distributed set of computer make decisions by consensus, handles failure, etc.
  • Flume and Scribe are logging services, Flume is an Apache project and Scribe is an open-source Facebook project. Both aim to make it easy to collect tons of logged data, analyze it, tail it, move it around and store it to a distributed store.
  • Google BigTable and it's open source twin HBase were meant to be read-write distributed databases, originally built for the Google Crawler that sit on top of GFS/HDFS and MapReduce/Hadoop. Google Research Publication: BigTable
  • Hive and Pig are abstractions on top of Hadoop designed to help analysis of tabular data stored in a distributed file system (think of excel sheets too big to store on one machine). They operate on top of a data warehouse, so the high level idea is to dump data once and analyze it by reading and processing it instead of updating cells and rows and columns individually much. Hive has a language similar to SQL while Pig is inspired by Google's Sawzall - Google Research Publication: Sawzall. You generally don't update a single cell in a table when processing it with Hive or Pig.
  • Hive and Pig turned out to be slow because they were built on Hadoop which optimizes for the volume of data moved around, not latency. To get around this, engineers bypassed and went straight to HDFS. They also threw in some memory and caching and this resulted in Google's Dremel (Dremel: Interactive Analysis of Web-Scale Datasets), F1 (F1 - The Fault-Tolerant Distributed RDBMS Supporting Google's Ad Business), Facebook's Presto (Presto | Distributed SQL Query Engine for Big Data), Apache Spark SQL (Page on apache.org ), Cloudera Impala (Cloudera Impala: Real-Time Queries in Apache Hadoop, For Real), Amazon's Redshift, etc. They all have slightly different semantics but are essentially meant to be programmer or analyst friendly abstractions to analyze tabular data stored in distributed data warehouses.
  • Mahout (Scalable machine learning and data mining) is a collection of machine learning libraries written in the MapReduce paradigm, specifically for Hadoop. Google has it's own internal version but they haven't published a paper on it as far as I know.
  • Oozie is a workflow scheduler. The oversimplified description would be that it's something that puts together a pipeline of the tools described above. For example, you can write an Oozie script that will scrape your production HBase data to a Hive warehouse nightly, then a Mahout script will train with this data. At the same time, you might use pig to pull in the test set into another file and when Mahout is done creating a model you can pass the testing data through the model and get results. You specify the dependency graph of these tasks through Oozie (I may be messing up terminology since I've never used Oozie but have used the Facebook equivalent).
  • Lucene is a bunch of search-related and NLP tools but it's core feature is being a search index and retrieval system. It takes data from a store like HBase and indexes it for fast retrieval from a search query. Solr uses Lucene under the hood to provide a convenient REST API for indexing and searching data. ElasticSearch is similar to Solr.
  • Sqoop is a command-line interface to back SQL data to a distributed warehouse. It's what you might use to snapshot and copy your database tables to a Hive warehouse every night.
  • Hue is a web-based GUI to a subset of the above tools - http://gethue.com 

Lets start from beginning: 



Tools and technologies for Bigdata:

Apache spark - Apache Spark is an open-source data analytics cluster computing framework originally developed in the AMPLab at UC Berkeley.[1] Spark fits into the Hadoop open-source community, building on top of the Hadoop Distributed File System (HDFS).[2] However, Spark is not tied to the two-stage MapReduce paradigm, and promises performance up to 100 times faster than Hadoop MapReduce for certain applications.

Database pipelining -
                      As you will notice it's just not about processing the data, but involves a lot of other components. Collection, storage, exploration, ML and visualization are critical to the proect's success.


SOLR -  Solr to build a highly scalable data analytics engine to enable customers to engage in lightning fast, real-time knowledge discovery.
        Solr (pronounced "solar") is an open source enterprise search platform from the Apache Lucene project. Its major features include full-text search, hit highlighting, faceted search, dynamic clustering, database integration, and rich document (e.g., Word, PDF) handling. Providing distributed search and index replication, Solr is highly scalable.[1] Solr is the most popular enterprise search engine.[2] Solr 4 adds NoSQL features

S3 - Amazon S3 is an online file storage web service offered by Amazon Web Services. Amazon S3 provides storage through web services interfaces. Wikipedia

Hadoop - Apache Hadoop is an open-source software framework for storage and large-scale processing of data-sets on clusters ofcommodity hardware. Hadoop is an Apache top-level project being built and used by a global community of contributors and users. It is licensed under the Apache License 2.0.  Apache Hadoop

MapReduce : Hadoop MapReduce is a software framework for easily writing applications which process vast amounts of data (multi-terabyte data-sets) in-parallel on large clusters (thousands of nodes) of commodity hardware in a reliable, fault-tolerant manner.



A MapReduce job usually splits the input data-set into independent chunks which are processed by the map tasks in a completely parallel manner. The framework sorts the outputs of the maps, which are then input to the reduce tasks. Typically both the input and the output of the job are stored in a file-system. The framework takes care of scheduling tasks, monitoring them and re-executes the failed tasks.

Corona :

Corona, a new scheduling framework that separates cluster resource management from job coordination.[1] Corona introduces a cluster managerwhose only purpose is to track the nodes in the cluster and the amount of free resources. A dedicated job tracker is created for each job, and can run either in the same process as the client (for small jobs) or as a separate process in the cluster (for large jobs).


One major difference from our previous Hadoop MapReduce implementation is that Corona uses push-based, rather than pull-based, scheduling. After the cluster manager receives resource requests from the job tracker, it pushes the resource grants back to the job tracker. Also, once the job tracker gets resource grants, it creates tasks and then pushes these tasks to the task trackers for running. There is no periodic heartbeat involved in this scheduling, so the scheduling latency is minimized. Ref : Under the Hood: Scheduling MapReduce jobs more efficiently with Corona


HBase : HBase is an open source, non-relational, distributed database modeled after Google's BigTable and written in Java. It is developed as part of Apache Software Foundation's Apache Hadoop project and runs on top of HDFS (Hadoop Distributed Filesystem), providing BigTable-like capabilities for Hadoop. That is, it provides a fault-tolerant way of storing large quantities ofsparse data (small amounts of information caught within a large collection of empty or unimportant data, such as finding the 50 largest items in a group of 2 billion records, or finding the non-zero items representing less than 0.1% of a huge collection).

Zookeeper - Apache ZooKeeper is a software project of the Apache Software Foundation, providing an open source distributed configuration service, synchronization service, and naming registry for large distributed systems.[clarification needed] ZooKeeper was a sub project of Hadoop but is now a top-level project in its own right.

Hive - Apache Hive is a data warehouse infrastructure built on top of Hadoop for providing data summarization, query, and analysis. While initially developed by Facebook, Apache Hive is now used and developed by other companies such asNetflix. Amazon maintains a software fork of Apache Hive that is included in Amazon Elastic MapReduce on Amazon Web Services.

Mahout - Apache Mahout is a project of the Apache Software Foundation to produce free implementations of distributed or otherwisescalable machine learning algorithms focused primarily in the areas of collaborative filtering, clustering and classification. Many of the implementations use the Apache Hadoop platform. Mahout also provides Java libraries for common maths operations (focused on linear algebra and statistics) and primitive Java collections. Mahout is a work in progress; the number of implemented algorithms has grown quickly,[3] but various algorithms are still missing.

Lucene is a bunch of search-related and NLP tools but it's core feature is being a search index and retrieval system. It takes data from a store like HBase and indexes it for fast retrieval from a search query. Solr uses Lucene under the hood to provide a convenient REST API for indexing and searching data. ElasticSearch is similar to Solr.

Sqoop is a command-line interface to back SQL data to a distributed warehouse. It's what you might use to snapshot and copy your database tables to a Hive warehouse every night.

Hue is a web-based GUI to a subset of the above tools. Hue aggregates the most common Apache Hadoop components into a single interface and targets the user experience. Its main goal is to have the users "just use" Hadoop without worrying about the underlying complexity or using a command line

Pregel and it's open source twin Giraph is a way to do graph algorithms on billions of nodes and trillions of edges over a cluster of machines. Notably, the MapReduce model is not well suited to graph processing so Hadoop/MapReduce are avoided in this model, but HDFS/GFS is still used as a data store.

NLTK - The Natural Language Toolkit, or more commonly NLTK, is a suite of libraries and programs for symbolic and statistical natural language processing (NLP) for the Python programming language. NLTK includes graphical demonstrations and sample data. It is accompanied by a book that explains the underlying concepts behind the language processing tasks supported by the toolkit, plus a cookbook.

NLTK is intended to support research and teaching in NLP or closely related areas, including empirical linguistics, cognitive science, artificial intelligence, information retrieval, and machine learning.

For Python-
                      Scikit Learn

                      Numpy

                      Scipy

Freebase - Freebase is a large collaborative knowledge base consisting of metadata composed mainly by its community members. It is an online collection of structured data harvested from many sources, including individual 'wiki' contributions.

DBPedia : DBpedia (from "DB" for "database") is a project aiming to extract structured content from the information created as part of theWikipedia project. This structured information is then made available on the World Wide Web. DBpedia allows users to query relationships and properties associated with Wikipedia resources, including links to other related datasets. DBpedia has been described by Tim Berners-Lee as one of the more famous parts of the decentralized Linked Data effort.

Visualization tool
                      ggplot in R
                      Tableu
                      Qlikview
                     
Mathematics : )

                      Calculus, Statistic, Probability, linear algebra and coordinate geometry

NER- Named Entity Recognition (NER) labels sequences of words in a text which are the names of things, such as person and company names, or gene and protein names.

Faceted search : Faceted search also called faceted navigation or faceted browsing, is a technique for accessing information organized according to a faceted classification system, allowing users to explore a collection of information by applying multiple filters. A faceted classification system classifies each information element along multiple explicit dimensions, called facets, enabling the classifications to be accessed and ordered in multiple ways rather than in a single, pre-determined, taxonomic order

Source : Wikipedia, the free encyclopedia

Thursday 7 August 2014

40 websites make you clever



             The indexed web contains an incredible 14 billion pages. But only a tiny fraction help you improve your brain power. Here’s 40 of the best.

1)khanacademy.org - Watch thousands of micro-lectures on topics ranging from history and medicine to chemistry and computer science.

2)freerice.com – Help end world hunger by correctly answering multiple-choice quizzes on a wide variety of subjects.

3)whizzpast.com - Learn about our awe inspiring past all in one wonderful place.Cleverwebsites4)artofmanliness.com - Blog/site dedicated to all things manly, great for learning life skills and good insights.

5)unplugthetv.com – Randomly selects an educational video for you to watch.

6)coursera.org – Educational site that works with universities to get their courses on the Internet, free for you to use.

7)mentalfloss.com – Interesting articles guaranteed to make you smile and get you thinking.

8)lizardpoint.com – A collection of browser-based games and the like to improve knowledge of geography, math, and such.10)ted.com - Collection of TED (Technology, Entertainment and Design) talks in which knowledgeable speakers address a variety of topics in short videos (< 18 minutes)

11)lifehacker.com – Learn to hack life! Tips and tricks for improving all areas of your life.

13)lifehackninja.com - Recently started life hack site that offers lists of useful knowledge

14)good.is/infographics – Collection of infographics which showcase useful information in an easy to digest visual format.

15)mathrun.net - Practice your basic maths skills with a simple game.

16)conversations.nokia.com – Learn about the world’s most innovative smartphones and future mobile tech.

17)Litro.co.uk – All you need to know about the wonderful world of books and writing.

18)zenhabits.net - Blog about improving your life by making it simpler.

19)justinguitar.com - Hundreds of free guitar lessons as well as some basic music theory.

20)duolingo.com – Learn a new language for free while helping to translate the web.

21)memrise.com - Learn things quickly and efficiently with flashcards that are spaced based on the spacing effect.

22)cookingforengineers.com - A site containing one man’s explorations in food, with step by step instructions for making a wide variety of dishes.

23)thedatingspecialist.com – A weekly blog run by a professional dating coach, offers advice on a wide variety of relationship topics.

24)nerdfitness.com – Fitness resource for the average person.openculture.com – Compendium of free learning resources, including courses, textbooks, and videos/films.

25)charitynavigator.org – Site dedicated to reviewing charities so you can easily research any you’re interested in.

26)lettersofnote.com – Their tagline says it all: “Correspondence deserving of a wider audience”

27)thedailymiscellany.com – A daily dose of useful knowledge, quotes, and other random things.

28)MIT Open Courseware - Free access to quite a few MIT courses that are on par with what you’d expect from MIT.

29)codecademy.com – Website packed with introductory courses for various programming languages and web technologies.

30)investopedia.com – Learn about the world of finance, from basic terminology to in depth analysis of various areas of investing.

31)udacity.com – Free courses mainly focused around mathematics and programming. Aimed at people of all skill levels.

32)lang-8.com – Write posts in a language you’re trying to learn, get them critiqued by a native speaker (and in turn help that native speaker learn your language)!

33)careerbuilder.com - Job hunting site to help you find a new career.

ureddit.com – Reddit’s very own University. Learn from redditors, or try to teach them.

34)engineerguy.com – Collection of videos in which Bill breaks down various feats of engineering in layman’s terms.

35)zooniverse.org – Take part in a huge variety of interesting studies of nature, science, and culture.

36)quora.com – Ask questions you’re curious about, answer those you’re knowledgeable about. Tailor your own feed to fit what interests you.

37)thenewboston.org - A fascinating collection of videos and tutorials related to maths and computer science.

38)aldaily.com - Aggregation of articles from various higher education journals and publications.

39)noexcuselist.com/everything – A huge list of awesome sites to learn from.

40)livemocha.com – Community dedicated to the idea of everyone being fluent in multiple languages, teach or be taught another language with the goal of conversational fluency.


Thursday 10 July 2014

10 Cool New Tech Ideas to Help You Market Your Business

Facial recognition, eavesdropping apps, augmented reality...these aren't security protocols but new ways for you to be able to market your business.

Marketing isn’t what it used to be.
Today we’ve got billboards that can ascertain our ages and genders, movie trailers that can change in real time depending on who the viewer is, and mobile apps that can detect what commercial is airing on TV and send viewers related coupons on their phones. And that’s just a start.
There’s no doubt technology is ushering in some fascinating new ways of connecting brands with consumers. In fact, to say technology is rapidly evolving doesn’t even fit anymore.
Disruptive is an adjective more likely to be used today because, honestly, tech exponentially builds on itself, resulting in a veritable sandstorm of innovation.
“There’s definitely a renaissance of marketing technology,” says Andrew Frank, research vice president in Gartner’s media industry group. The Web and the digital revolution, in general, along with mobile have unleashed a lot of things…There’s certainly a lot to keep track of.”
Here is our roundup of 10 of the latest developments in technology that can impact your marketing.
1. Facial RecognitionGiving away free product is a great way to build your brand, right?
Unilever, the company that makes things like Lipton soup and Suave shampoo also makes ice cream and figured it would use some of it to get customers to smile—literally. It worked with the high-tech marketing company SapientNitro in London to make a vending machinecalled “Share Happy” that uses facial recognition to give people ice cream bars for smiling.
They even designed the box so people can upload their smiling mugs to Facebook, pushing the brand even farther out into the consumer space. When the machine debuted in Lisbon last year, it vended 52,000 ice cream bars in two days with no media support or advertising. That’s about 600 percent more than a regular ice cream machine.
The award-winning box is now on world tour doling out brand recognition as it goes.
“Facial Recognition is being used not only on experiential campaigns but even on digital signage,” says Rob Gonda, global head of emerging technology and innovation for SapientNitro. “For instance, a lot of our installations…detect things like gender, age bracket and start serving more targeted advertising on digital signage and digital displays.”
What’s different today compared to a few years ago, Gonda says, is brands have to be situationally aware and contextually relevant.
“We have evolved from broadcasting messages to [delivering] more digitally precise messages. But actually understanding the context of the situation or the conversation…increases the likelihood of better engagement [between] brands [and] people,” he says.
Gonda says SapientNitro is currently working on interactive displays for retail spaces that use a mashup of rear-projected video, touch, gesture recognition and facial recognition.
And SapientNitro is far from the only company doing it.
According to The Los Angeles Times, The Venetian Resort in Las Vegas is using facial recognition to tailor restaurant and entertainment ads for passersby and big brands like Adidas are installing and testing digital walls with facial recognition in a handful of stores to display age-appropriate shoes to customers.
2. Hyper Targeting
While targeted advertising has been around for a while, today the big thing is to make ads super specific.
When people spend time online—surfing, shopping or commenting on social networks—they leave a digital footprint that huge data mining systems can track. That means advertisers can profile nearly everybody today in terms of preferences, habits and demographics.
Cambridge, Massachusetts-based ChoiceStream has an ad platform called CRUNCH that uses predictive intelligence—the same technology retailers like Zappos use to suggest shoes to customers. It selects consumers to target based on “hyper-segments” – a high performing cross-section of audiences with two or more attributes applied, meaning ads are delivered to people who are most likely to respond to them. 
To illustrate, ChoiceStream COO Eric Bosco says one of the company’s clients is an antivirus software provider and to hyper-target ads for this particular customer, CRUNCH not only looks for people shopping for software but also for those who order pizza online and get online legal self help. The results from that narrowly focused segment end up being five times better than when the company only uses one attribute.
“That type of activity is a proxy for how active these users are on the Internet. They’re probably consuming a lot more online content, and that exposure has the side effect of exposing them to more viruses and risks, and so they’re much more likely to need software to clean their machines,” he says.
The company’s advertising is so highly targeted; in fact, that it has actually has a hard time getting business from advertisers who market things everyone needs.
“We were recently in Minneapolis talking with an ad agency there and their main issue with CRUNCH is that they place ads for butter. Everybody needs butter, so how do you target butter? We excel when things are a little more focused than that,” Bosco says.
3. Eavesdropping Apps
While they’re still mostly in concept stage, Gartner’s Andrew Frank says companies are starting to play around with advertising through new apps that can tell what show you’re watching on TV or listening to on the radio.
He says Yahoo bought a company called IntoNow, which has a fingerprinting technology that can listen to and identify what’s on TV, and let people check in to a show or ad.  “Pepsi Max ran a promotion where if you tag their ad when it runs using this app you can get a coupon for some free Pepsi Max,” he says.
“The media synchronization that can recognize the audio track of a TV show can also be used on radio so if you’ve got a mobile phone and there’s a radio ad for example, you might be able to use that to distribute a coupon directly from the audio, sort of like an audio QR code directly from the ad to the mobile phone through the microphone,” he says.
4. Augmented Reality
Swiss watchmaker Tissot, used Augmented Reality (AR) outside of Harrods in London to let people “try on” its watches—virtually. They handed out black and white paper wristbands and when someone stood in front of the company’s large LCD screen, a webcam behind it sent the coordinates of the paper watch into an AR system which then displayed video of the person on the screen wearing a 3D virtual version of the watch. The person could even play around with the watch’s touch screen features.
Tissot now uses the high-tech display at events in the States, as well, such as at the MotoGP motorcycling championship at the Laguna Seca raceway in California, where the company is the official timekeeper.
But the space where you’ll be seeing more AR is mobile.
According to this infographic created by the U.K.-based AR developer Hidden Creative, AR is now being used for a wide range of marketing activities—in everything from riveting displays at conferences and for sales pitches to loyalty programs at the Point of Sale. Hidden says 35 AR apps are launched every month.
And some of them look pretty amazing.
Junaio has iOS and Android platforms that brands can use to customize AR for their customers’ needs.
Valpak, for example, has a channel on the Junaio app that launches the smartphone’s camera and GPS to overlay a set of 3D icons in real time showing the available Valpak coupons in the vicinity as the user scans the surroundings, with a radius varying from five feet to 20 miles away.
5. Mobile: Codes and Spot Targeting
Quick Response (QR) codes are the little black-and-white pixilated squares commonly found in print ads, on products, displayed on TV, and even stitched into fabric. If you have a phone with a camera and the right kind of reader app, you can scan a QR code to display text, contact information, or open a Web page.
People are also scanning UPC bar codes—especially the ones found on products you’d find at a grocery or drug store—for product pricing and information.
According to David Javitch, VP of marketing for New York City-based Scanbuy, the company has seen an 800 percent increase since a year ago in the number of barcode or quick response (QR) code scans it processes and is now doing more than one per second. 
While Scanbuy currently has 8 million people using its ScanLife app on their phones and has worked with big companies such as The Home Depot and Taco Bell on huge code campaigns, the company is rolling out a platform that will allow regular people and small businesses to create their own QR codes starting at $25 a month.
While pricier than QR codes, StarStar Numbers might be a titch faster from a consumer perspective since it only involves one step. They’re branded vanity mobile numbers that let consumers call a brand’s name. For instance, if you’ve got a craving for Samoas or Thin Mints, you can call (not text) **GSCOOKIES and you’ll instantly receive a text message with a link to a Girl Scout Cookie Locator app.
In addition to smart phones giving consumers near-instant access to marketing or product information and digital content, they also are empowering businesses to do local spot targeting that only serves ads to people near a certain geographic location. For instance, a movie theatre could broadcast show times to only those smart phone users within a mile radius who happen to be doing things like playing Angry Birds, listening to Slacker Radio, or other similar activities.
Millennial Media, a Baltimore-based mobile ad platform, delivers such ads to consumers who are browsing the mobile Web or using apps. The company says advertisers can use a variety of formats such as banner ads, rich media ads or interstitials.
6. VideoGartner’s Andrew Frank says the video space is seeing some interesting innovation, particularly in the area of what he calls “dynamic creative” that lets a company or advertiser change video content in real time.
“For example, you might re-edit a movie trailer on the fly depending on whether the viewer is male or female or in the North or in the South,” he says, holding up Tumri and Teracent as examples of ad platforms that can do it.
Cloud-based Flite is another ad platform that allows for real-time editing of ads. It can incorporate real-time content streams such as videos, polls, and other interactive elements into an ad, and even pull in social media content from Facebook, Twitter and YouTube—all of which can be switched up at any time. 
For instance, a restaurant having a slow couple of hours could update its display ad with a flash deal that includes a poll and its Twitter stream that mentions the promotion.
Flite's pricing is about $1 CPM (cost per thousand views) but it depends on what you set your ad up to do. Simpler ones are cheaper but ads with customization such as sweepstakes or video galleries are more.
But what if putting things like YouTube in your ads sounds like a great idea, but you don’t have any content?
Near Networks just launched an online video channel platform for local businesses across the country. For $1,499 it will send a professional camera crew to your location and create a 90 to 120 second high definition micro-documentary for your business. 
Once the video is done, Near Networks helps its clients create and manage a video channel on YouTube that can be embedded on a Web site, blog or on social media outlets.
7. Incentives and Virtual Currency
Gartner’s Andrew Frank says incentivized ads in games such as Farmville, Mafia Wars and Angry Birds tap into a huge audience of social gamers who play every day. “It’s the idea of giving people game credits or virtual goods or other things that they can use in a game in exchange for watching ads or interacting with them or otherwise engaging with the sponsor’s content,” he says.
 “Virtual currency,” is what Chicago-based Lab42 calls it.
The year-old start-up not only creates some of the great infographics showing up lately on media outlets such as Mashable or HuffPost Tech, but it also offers an online market research tool that uses social networks to do market research. 
You can either create your own survey using the Lab42 dashboard or they’ll do it for you. After selecting your target market, they launch the survey across various social networks and social media partner sites that offer virtual currency in exchange for the respondents’ time and opinion. By reaching out to the desired audience of their client’s target market, the Lab42 research team is able to return results in just a few days, with packages that start at just $500.
Another example: Peer-to-peer promotion.
Calyp is a network that rewards consumers with cash, gifts and discounts for promoting products they like with their friends and networks.
Using either the Calyp mobile app or Calyp.com, an endorser can broadcast a Calyp link—through a text message, blog post or social status update—that features her favorite line of clothing, for example, and let people know there’s a big sale going on. When her friends or followers click on the link, she gets the credit.
Getting your business into the Calyp network will run anywhere from $950 to $25,000 a month, so it’s not cheap. Blue Calypso, the company behind Calyp, says the investment pays off.
Blue Calypso says it has launched more than 400 campaigns, reaching more than 1.15 million mobile and social media consumers with a 325 percent click-through rate and 21 percent intent to purchase.
“We’re not a good fit for a new business,” says Sabrina Dee, Blue Calypso’s CMO. “It’s about using your loyal customers that love you and can speak with an education about your brand because they use your brand—sharing with people that are like-minded about why they would also like it.”
8. Social Analytics
Today there are great analytics tools that help brands engage with customers and potential customers on social networks. Here’s a slide show of the best social CRM programs that features several good ones.
“I believe that one of the most important trends in the market is the increasing participation of consumers/customers on social networks, and the 'big data' problem - and opportunity - it is creating for marketers who can effective analyze and act on that data,” says Chris Selland, CMO for social marketing platform Terametric and expert on the Focus network.
“Traditionally, marketers have focused on analyzing data that is theirs - visitors to our Web site, callers to our call center. But social data lives on social networks - outside the walls of the organization…It does not belong to the company, but the ability to analyze and act effectively on it is key,” he says.
9. Web
Is your Web site doing the best job it can for you?
Plenty of great tools now exist for getting measurable data about the effectiveness of yoursite design and how well you’re marketing to your traffic.
EyeTrackShop is different. It’s a Sweden-based digital marketing research firm with offices in New York and Los Angeles that conducts eye-tracking studies to gauge the effectiveness of not only Web site pages but also online ads, banner ads, print ads and packing designs to see if your audience will notice what you want them to. Some of their clients include P&G, Google, SnapFish and Clorox.
You simply send your sample design to EyeTrackShop, along with the demographic you want to test. From there, EyeTrackShop enlists its U.S. database of 6 million users who use their own Web cams to send video to EyeTrackShop, which then uses retina scanners to gauge gaze duration and location. 
EyeTrackShop says it returns results in 48 hours, charges $2,500 for one stimuli test and $1,800 for two tests.
Another useful tool you might check out is Gigya, which you can use on your Web site to allow users to login using their preferred social identity.
While you can use something like Facebook for Websites to let people log on using Facebook, if you want to save the development time, Gigya offers what it calls a "Super API" which lets users to log in to your site with over 25 different identity providers such as Facebook, Google, Twitter, AOL, Yahoo and more. The company says you’d be surprised how many people choose to log in with non-Facebook identities.
Gigya also offers a suite of social plugins that can to your site features such as comments, sharing, ratings and reviews, games and chat.  
When users log in to your site using their social identities, they grant access to data inside their social apps which Gigya can then collect and store on your behalf. For marketing purposes, this can allow you to not only serve them specific content, but targeted ads as well.
10. Deals
Daily Deal sites are everywhere and whether using them is good for business is a topic Inc. has written about elsewhere.
Signpost, however, does it differently than most. Instead of taking a share of each deal—money that should be going into your pocket—it lets you set up as many deals as you want on its platform and only charges you $99 a month (with the first month free).
Its self-service platform gives merchants complete control over the discount amount, when
each deal is announced, the timeframe in which the deals can be redeemed, and the limit of how many deals can be sold.
Signpost markets the deals businesses create through its Web site, daily newsletter and social media following.