SECTION used in recommendation system.1) Content Based Recommendation

SECTION A:Title, Technical Area(s) addressed, Period of Performance, Estimated Cost, Name/Address of Company, Technical and Contracting Points of Contact (phone, fax and email)SECTION B: Task ObjectiveAnalytics for Situation AwarenessMachine Learning Section C: Technical Summary and Proposed Deliverables.  As entrepreneurs we are constantly looking for better ways to serve our clients. When software analysis & prediction plays a significant role in the equation, we need to take a closer look at the advantages computer applications have in reaching this goal.Let us go through the few examples mentioned above and look at the possibilities of making the software smarter – smart enough to play the role of an assistant possessing more than simple rules.Planning – Although the human planner can still plan the personnel for complex and rule-ridden time shifts (which can really get hairy with thousands of employees), now the application can recommend perfectly good assignments. It can learn from the planner’s past plans, as well as the past data of employee time preferences. The application assists the planner rather than take over his job. The planner can overrule, alter, or just approve the automatically generated plan. His or her changes will affect the future recommendations of the system. Recommendation system is one of the application of machine learning. When we want to recommend something to a user, the most logical thing to do is to find people with similar interests, analyze their behavior, and recommend our user the same items. Or we can look at the items similar to ones which the user bought earlier, and recommend products which are like them.Following are some machine learning models used in recommendation system.1)   Content Based Recommendation (CBR): CBR provides personalized recommendation by matching user’s interests with description and attributes of items. For CBR, we can use standard ML techniques like Logistic Regression, SVM, Decision tree etc. based on user and item features for making predictions for eg: extent of like or dislike. Then, we can easily convert the result to ranked recommendation.2)   Collaborative filtering (CF)i) user-based collaborative filteringii) item-based collaborative filtering3)   Matrix Decompositions4)   Clustering5)   Deep Learning Approach The use of Machine Learning in recommendation solely depend upon the objective and reformulating the problem into best suited Machine Learning algorithm. Transaction monitoring – The application of Machine Learning techniques on transaction monitoring may take several forms. We could shortly mention the ability of these types of algorithms to detect anomalies alleviating the need to specifically code threshold-based rules and also the capability of Machine Learning techniques to learn from the experts and generalize their strategies. When integrated in a transaction monitoring module, the system can alert the expert to pay her attention to outliers or serve as an early warning system, allowing her to watch their development.Web Sites – If taken to the next step, online applications can learn the flows and needs from each user. How does he use the application? What parts of it need to become highly accessible for this specific user? Making sense of large number of users makes it possible for the application to recognize different kinds of usages users make of the application and assist them by predicting their needs.