Rohit Kumar's Projects
The dataset is picked from the Open Baltimore website. The Dataset presents the 911 police calls made for different service at different Date and Time, from different location of Baltimore marking the priority of call from low to high and Emergency or Non-Emergency calls.
Image indexing, Image captioning and Image retrieval is a hot topic of research in NLP and Computer Vision now-a-days, but the work on it have started decades ago. The context of the image captions matters a lot while talking of image processing. Presence of a name/text in a caption related to a given context provides useful information concerned with the image like who is associated with given image. This is the Graduate Paper work as a part of NLP Coursework at UMBC. This paper discusses the related works and findings in the field of text and image captioning and modeling using the results and findings from the first five paper mentioned in references section of this paper.
The Dataset for Breast Cancer is picked from sklearn and is splitted into equal halves for Train and Test Dataset. I have used Logistic Regression and Support Vector Machines with linear and RBF kernels for making the analysis. Further, I have made use of scikit's StandardScaler to standardize the data. Using the unscaled data, I have tuned the parameters of each model using GridSearchCV. For Logistic Regression and SVC models, I have tuned the C parameter and also tuned the gamma parameter for the RBF kernel. Atleats, discussed about the coefficients of the LogisticRegression and Linear Support Vector Machine models and explained what is their significance corresponding to the features.
Implementation of Chord in Go
Implementation of Chord DHT(Distributed Hash Table) paper
Create an API Server using Node.js, Express and JSON file which have the following services. 1) Add a user (HTTP Post), 2) Update a User (HTTP Post), 3) Delete a User (HTTP delete), 4) Get all users (HTTP Get). Get a specific user by passing the user ID (HTTP Get) Use a file on the node server to maintain user information.
Cheat Sheets
TED is a non-profit American media organization, spreading ideas, usually in the form of short and powerful talks (18-30 min video) covering areas from science to business to education to global issues. Most of these talks are available in more than 100 other languages. In this project, we have taken 3 different datasets for the available TED talks. These datasets include details about ~2500 TED talks, where one of the datasets provides the transcript of each TED Talk topic and second one provides the details like length of talks, views, release date, languages, rating, tags, views, comments etc. The third dataset provides speaker’s details which includes speaker’s career field, their background detail, etc. We analyzed the available datasets to get useful inferences to conclude how gender, duration, languages, views, comments, speaker’s occupation, rating, topic of TED talk and the easiness of following the speech affects the popularity of the talk.
Deep Learning Specialization by Andrew Ng on Coursera.
Golang implementation of the Chord protocol
Chord implementation with DHT in Golang
MOA is an open source framework for Big Data stream mining. It includes a collection of machine learning algorithms (classification, regression, clustering, outlier detection, concept drift detection and recommender systems) and tools for evaluation.
My Python n-gram Language Model from an NLP course. Since there are so public implementations, I feel free to post mine.
SVM Solver in Python (http://www.cs.huji.ac.il/~shais/papers/ShalevSiSrCo10.pdf)
100+ Python challenging programming exercises
Web Dev workshop
How to query your API using Vuex in your Vue application. Use Axios to interact with the API created using Vue. Axios is Client Side network request library to send an HTTP Request. Axios and Vue-Axios library makes API calls super simple. Actions are meant to be asynchronous while mutations should happen as near to instantly as possible.