rajpatel0909 Goto Github PK

data-cleaning-and-munging

Convert, extract, repurpose, transform, document and learn data

data-clients-and-information-servers

Understand and implement Application Programming Interface (API) based programmatic data collection from popular (/public) data sources (Data clients).  Process the data collected for extracting basic information.  Serve the information extracted through simple applications and visualization (Information servers).

datingpgm

Bayesian Model to infer dating habits using an experimental dataset. The model helps us infer various features that affects the decision making of the host and partner. We compare exact inference and approximate inference methodologies to infer probability queries, marginal queries and MPE.

drone

DRONE DEPLOY

generative-chatbot

Chat-bot to answer the domain specific questions or to talk in human like way Chatbots are also called Conversational Agents or Dialog Systems. The Chatbot implemented in this paper is generative based. The basic algorithm used for developing Chatbot in this paper is Long Short Term Memory which a version of Recurrent Neural Network.

handwritten-digit-recognition

The project is about using well-known technique Neural Network of machine learning, for developing an algorithm such that program can learn by itself about recognizing handwritten digit on feeding training data to it. Feedforward and Backpropagation is used to compute gradient for cost function and cost function is also regularized. I used octave for this project.

language-identification

In natural language processing, language identification is the problem of determining which natural language given content is in. Computational approaches to this problem view it as a special case of text categorization, solved with various statistical methods. The data set is lines of movie subtitles from 21 different languages. The task here is to train a model based on training data set, learn the hyper parameters on validation set and identify the language of subtitles in unknown languages. The model implemented for this task is Naïve Bayes model. Each line in training data set is considered as a single document. All documents are represented as Bag of n-grams.

lipreading

Lip Reading

machine-learning

Linear Regression, Classification, Neural Network and Handwritten Digit Recognition Project.

ml-project-2

description is in pdf

mouthdetection

Mouth Detection for HMM purpose

mysky

question-answering-search-engine

Question Answering based Search Engine. Read Report for more details

stereo-vision--image-segmentation-and-image-search-engine

Developed depth map of image by disparity estimation using block matching and dynamic programing and synthesized new view using disparity maps.  Segmented various objects in image using Mean-Shift Segmentation.  Developed Content Based Image Retrieval System (Image Search Engine) using OpenCV.

tdg

tower-defense-game

The project is a 3d game which is a unique version of tower defense game, developed using unity game engine and C Sharp language. The game development is platform independent and can be available for multiple platforms like windows (PC and mobile), android, IOS, mac, etc. The game is about defending and/or offending tower or similar structures based on strategies and artificial intelligence.

twitter-project

The project mainly includes four tasks: access the twitter Application Programming Interface(API) using python to collect tweets as data, estimate the public's perception (the sentiment) of a particular term or phrase, analyze the relationship between location and mood based on a sample of twitter data, determine the sentiment of other terms or phrases on basis on analysis.