ML Algorithms

All the snippets need to be optimized to be made scalable and are only initial implementations of the standard ML algos

Dataset Sources

* Air Passenger Dataset

* [Power Consumption Patterns ] - Not Available ONLINE

* Subject Classification Dataset

* Geo-location Coordinates Training Dataset

* Geo-location Coordinates Test Dataset

* Sleep Deprivation Dataset

* Rock Mines classification dataset

This Week TODO:

• UNIVARIATE and MULTIVARIATE LINEAR REGRESSION(W/WO GRADIENT DESCENT) -- DONE
• LOGISTIC REGRESSION on first 10 principal components of a dataset to CLASSIFY an Experiemnt as of Physics OR Chemistry -- DONE
• NEURAL NETS for binary geo-coordinates classification using back propogation -- DONE ** The Architecture of the neural net implemented is shown below ,with two hidden layers and two units in each hidden layer, apart from bias term **
• SUPPORT VECTOR MACHINES -- 3 dimensional feature set mapped to 180 dimensional feature set where m= #landmarks/#training egs using gaussian kernel
• K-Means Clustering --DONE
• Principal Component Analysis (PCA) --DONE
• Col Filtering --TODO

README will be updated ** very very frequently