Chinmay Wyawahare's repositories
Stock-Market-Sentiment-Analysis
Identification of trends in the stock prices of a company by performing fundamental analysis of the company. News articles were provided as training data-sets to the model which classified the articles as positive or neutral. Sentiment score was computed by calculating the difference between positive and negative words present in the news article. Comparisons were made between the actual stock prices and the sentiment scores. Naive Bayes, OneR and Random Forest algorithms were used to observe the results of the model using Weka
Denoise-Noisy-Docs
Removal of stains from noisy docs using image processing, machine learning, neural nets and autoencoder
NYCOpenData-Profiling-Analysis
Open Data Profiling, Quality and Analysis on NYC OpenData dataset with semantic profiling using fuzzy ratio, Levenshtein distance and regex
Foursquare-clone
Oingo is a new mobile app named that allows users to share useful information via their mobile devices based on social, geographic, temporal, and keyword constraints. The main idea in oingo is that users can publish information in the form of short notes, and then link these notes to certain locations and certain times. Other users can then receive these notes based on their own location, the current time, and based on what type of messages they want to receive
Twitter-clone
Webber is a distributed twitter like application written in Go using Raft consensus algorithm for leader election, log replication, handle node failure mechanism. The application uses etcd Raft library and has 3 microservices which communicate using gRPCs
Aiddata
AidData's Core Research Release 3.1 is a corrected snapshot of AidData's entire project-level database from April 2016. This database includes commitment information for over 1.5 million development finance activities funded between 1947 and 2013, covers 96 donors, and includes ODA, OOF flows, Equity Investments, and Export Credits where available.
Optical-Flow
Implementation of Lucas-Kanade and Horn-Schunck methods for optical flow
Reinforcement-Learning-Comparative-Study
Comparative study of Reinforcement Learning Algorithms on Ping Pong game: In the current design of experience replay we sample uniformly to obtain the minibatch and update the model. Devising a way to sample more experience points close to the tricky areas would help solving this problem, better the training rate and improve convergence. We designed a game environment for the Android platform as few such environments are available at the moment. Moreover, during the pre-processing game we removed the background and score to reduce clutter and increase likeliness of successful training. It would be interesting to see how restoring the background affects agent’s performance. Overall, our results show the capacity of Deep neural networks and how a generic reinforcement learning setup such as this could learn and play the game with very minimal domain knowledge.
Premier-League
Premier League analysis of seasons 2006/2007 to 2017/2018
SF-Opioid-Crisis
San Francisco (SF) has a long history of pushing the envelope on progressive public health solutions, including medical cannabis and needle exchange, before either was legal or broadly embraced. It is so out of proportion, that California passed a bill allowing SF to open Safe Injection Sites (SIS).
20Newsgroups-data-mining
Using LIME and Spark MLlib, validate sentiment analysis using two classes - Atheism and Christianity with spark data pipeline and calculating F1-score, accuracy and class probabilities
Data-Science-projects
Applying machine learning algorithms on various datasets
Denoising-Dirty-Documents-using-R
Removal of stains from documents by providing dirty and clean images as data-sets for the model to learn and the model would clean the test images by observing results from the training dataset using Linear Regression.
Fashion-MNIST
Fashion-MNIST is a dataset of Zalando's article images—consisting of a training set of 60,000 examples and a test set of 10,000 examples. Each example is a 28x28 grayscale image, associated with a label from 10 classes. Zalando intends Fashion-MNIST to serve as a direct drop-in replacement for the original MNIST dataset for benchmarking machine learning algorithms. It shares the same image size and structure of training and testing splits.
gandalf1819.github.io
Personal website
Histogram-Equalizer
Histogram equalization is a technique for adjusting image intensities to enhance contrast
Image-denoising
Image denoising techniques in Computer Vision using Box, Gaussian and Median filter
NYC-Parking-Violations
Using MapReduce and Spark to explore NYC Parking Violations
PhotoAlbum
PhotoAlbum is a photo album web application which allows searching using natural language through both text and voice using Lex, ElasticSearch, and Rekognition for an intelligent search layer to query your photos for people, objects, actions, landmarks and more.
photography
My online photography portfolio
Risk-Factor-Identification-using-Truck-Fleet-Sensor-Data
Computed truck mileage, driver risk factor using Hive and Pig to understand the risk the company is under from fatigue of drivers and over-used trucks and visualized the sensor data using Tableau to observe the impact of the factors on driver’s performance
Twitter-feed-analysis
Twitter Feed Analysis over twitter data to find most influential people, time zones when majority of users are available and the most common hashtags used on Twitter using Hive