Today professor  talked about recommendation system. First he showed the solution program for last assignment.  We can create an index and match information in it. 

Machine learning 

A machine learning system is an algorithm that is capable to learn from data. A ML system is said to learn from experience with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E. 

Machine learning is often dividend into two broad categories:

supervised Machine Learning: learns from existing data(experience) and is able to predict outcome(perform tasks) from new data

unsupervised machine learning 

Then, professor talked about the example for training phase. Machine could identify the picture of cats through machine learning.

About experience E, it can be input dataset, examples, features, vectors etc. To each example, a label or target is associate. In classification, label may be in a finite set and a real number in regression or prediction.

About the task, T, learning itself is not a task. Its common types are classification, transcription, regression

In classification, ML system is required to specify which of k categories a set of features x belong to. ML system produces a function y=f(x) that maps x to the one of the k categories. In transcription, ML system transcribes symbols from after observing unstructured forms.

Then we moved to next topic, recap. Studied definitions of Machine Learning and concepts associated with it together with examples. Learned Experience, Task, and performance Measure in the definition of Machine Learning. Reviewed examples and identified Experience, Task, and Performance Measure in them.

Feature Engineering

Feature Engineering is definition of features by transforming raw data based upon understanding of the domain. These features are input to machine learning system as experience.