A Recommender System for Amazon products based on Sentiment analysis of Reviews. Supervised Learning/Natural Language Processing/Collaborative filtering
The project goes through different techniques of supervised learning like AdaBoost, Random Forest, Logistic regression etc. to classify the underlying sentiment of the reviews given to every product and based on those sentiments recommendations are made to the users with the help of collaborative filtering.
Dec 2018.
[code]

Finding Customer segments for a whole sale distributor for delivery schedules: Unsupervised Learning(Clustering)
This project uses K-Means clustering, Gaussian Mixture model clustering,feature selection methods, Decision Trees, PCA, data cleaning and data wrangling techniques.
Mar 2018.
[code]

Predicting salary of a person from US Census data for Charity: Supervised Learning(Classification)
This Project uses SVM,Perceptron,AdaBoost classifier. Data wrangling and Data Cleaning is done from scratch.
Jan 2018.
[code]

Predicting potential clients who would open a long term deposit account at a Lisbon bank: Supervised Learning(Classification)
Uses generative probability models like SMOTE,and classifiers like neural networks,SVM,Minimum risk Bayes classifier and perceptron.
Nov 2017.
[code]

Predicting Boston house prices based on Historical data: Supervised Learning(Regression)
This project consists of regressors like Decision Tree Regressor and AdaBoost regressor.
Aug 2017.
[code]

Prediction of the underlying sentiments of Amazon book reviews: Supervised Learning(Multiclass Classification)
Uses Naive Bayes model to predict these sentiments after converting the data-set in bag of word format.
Feb 2017.
[code]

Exploratory Data Analysis of survival rates of titanic ship wreck.
Uses Decision tress and adaboost.
Nov 2016.
[code]