vikramjitsinghrathee/machine-learning-and-deep-learning-projects

▪Deep Learning: Convolutional Neural Network (CNN), Deep Neural Network (DNN) regression ▪Supervised Learning: Decision Tree, Support Vector Machine and Gaussian Naive Bayes ▪Unsupervised Learning: Principal Component Analysis (PCA) and K-means clustering ▪Reinforcement Learning: DDPG (Deep Deterministic Policy Gradient) algorithm

Short Description of the projects

Deep Learning/Dog Breed Classifier_CNN

In this project, given an image of a dog, the Convolutional Neural Network (CNN) will identify an estimate of the canine’s breed. If supplied an image of a human, the code will identify the resembling dog breed. 3 CNN architectures were analyzed, one built from scratch and 2 pre-trained (Resnet-50 and VGG-16). Also implemented Transfer learning with pre-trained CNN architecture (Resnet-50).

Supervised Learning/Finding Donors for CharityML

In this project, 3 supervised learning algorithms (Decision Tree Classification, Support Vector Machine Classification, Gaussian Naive Bayes Classification) were implemented to accurately model individuals' income using data collected from the 1994 U.S. Census. The goal with this implementation is to construct a model that accurately predicts whether an individual makes more than $50,000. This sort of task can arise in a non-profit setting, where organizations survive on donations.

Unsupervised Learning/Creating Customer Segments

In this project, dataset containing data on various customers' annual spending amounts (reported in monetary units) of diverse product categories for internal structure was analysed. The goal of this project is to best describe the variation in the different types of customers that a wholesale distributor interacts with. Doing so would equip the distributor with insight into how to best structure their delivery service to meet the needs of each customer. PCA (dimensionality reduction) and K-means clustering (unsupervised clustering algorithm) were utilized.

Reinforcement Learning/Teach a Quadcopter How to Fly

In this project Reinforcement Learning was utilized to train a quadcopter agent to fly using DDPG (Deep Deterministic Policy Gradient) algorithm.

Machine Learning Capstone/New York City Taxi Fare Prediction - Kaggle Competition_DEEP NEURAL NETWORK

The project deals with accurate estimation of the taxi fare given pickup/dropoff latitude and longitude, number of passengers, pickup date and time. The training and testing data was obtained from Kaggle (https://www.kaggle.com/c/new-york-city-taxi-fare-prediction). The Deep Neural NetworK (DNN) regression was performed for obtaining taxi fare predictions.

Model Evaluation and Validation/Predicting Boston Housing Prices

Utilized Boston Housing dataset which contains aggregated data on various features for houses in Greater Boston communities, including the median value of homes for each of those areas. Decision Tree Regressor was used to train on the housing dataset and then used to estimate the best selling price for your client homes.