Pinned Repositories
AutoEncoder
Data-Science--Cheat-Sheet
Cheat Sheets
Distributed-AI-on-the-Edge-Platform
Built a Distributed AI on the Edge Platform that is capable of managing & deploying tensorflow models using tensorflow serving. The platform was built using micro-service based architecture with features like auto-scaling, scheduling, centralized logging, monitoring, notification, high-availability and fault tolerance.
InodeBasedFileSystem
Developed inode based virtual file system on top of the Linux file system. Implemented all the operations which are basically supported by Linux file system like creating disk, mounting disk, unmounting disk, create file, open, delete, close file, list of files, read mode, write mode and append mode.
IRE-hashtag-generation
Designed and implemented a multi-modal hashtag generation pipeline that suggests new hashtags for an Instagram post with images & text. The approach used CNN for image classification and glove-embeddings for recommending semantically similar hashtags.
LinearHashing_B-Tree
LinuxFileExplorer
Developed terminal based file explorer which supports both Normal mode as well as command mode. Implemented all commands like cut, copy, delete, rename, goto, search, snapshot for both files and directories. Worked with data structures to handle files and its information. It also supports run time modifications of files.
Restaurant_Rating_Prediction
Implemented a model that is capable of predicting a restaurant rating taking into account several factors such as reviews and restaurant facilities. Analysis of review is done based on NLP techniques that include polarity analysis, TF-IDF which are all followed by pre-processing.
StackOverflow-Clone
Re-implemented Stackoverflow web based application in python programming with flask framework and MVC architecture. Worked with MYSQL database and sqlalchemy ORM model to handle data. It supports functionalities like ask question, post answer, upvote, downvote, views, answer later, post comment, search.
Wikipedia_Search_Engine
Implemented a search engine on the wikipedia dump of size 73.4 GB. In order to retrieve result faster and relevant, indexing and ranking is implemented. Relevance ranking algorithm is implemented using TF-IDF score to rank documents. Creating index takes around 14 hr on a given wikipedia dump. Result is retrieved in less than 1 second.
VatsalSoni301's Repositories
VatsalSoni301/Distributed-AI-on-the-Edge-Platform
Built a Distributed AI on the Edge Platform that is capable of managing & deploying tensorflow models using tensorflow serving. The platform was built using micro-service based architecture with features like auto-scaling, scheduling, centralized logging, monitoring, notification, high-availability and fault tolerance.
VatsalSoni301/AutoEncoder
VatsalSoni301/Data-Science--Cheat-Sheet
Cheat Sheets
VatsalSoni301/InodeBasedFileSystem
Developed inode based virtual file system on top of the Linux file system. Implemented all the operations which are basically supported by Linux file system like creating disk, mounting disk, unmounting disk, create file, open, delete, close file, list of files, read mode, write mode and append mode.
VatsalSoni301/LinearHashing_B-Tree
VatsalSoni301/Restaurant_Rating_Prediction
Implemented a model that is capable of predicting a restaurant rating taking into account several factors such as reviews and restaurant facilities. Analysis of review is done based on NLP techniques that include polarity analysis, TF-IDF which are all followed by pre-processing.
VatsalSoni301/UndoLoggingAndRecovery
VatsalSoni301/IRE-hashtag-generation
Designed and implemented a multi-modal hashtag generation pipeline that suggests new hashtags for an Instagram post with images & text. The approach used CNN for image classification and glove-embeddings for recommending semantically similar hashtags.
VatsalSoni301/Wikipedia_Search_Engine
Implemented a search engine on the wikipedia dump of size 73.4 GB. In order to retrieve result faster and relevant, indexing and ranking is implemented. Relevance ranking algorithm is implemented using TF-IDF score to rank documents. Creating index takes around 14 hr on a given wikipedia dump. Result is retrieved in less than 1 second.
VatsalSoni301/Classifier_Regression
VatsalSoni301/CNN_Tensorflow
VatsalSoni301/DecisionTreeImplementation
VatsalSoni301/Hackathon_1
VatsalSoni301/Hackathon_1_TensorflowServing_Docker
VatsalSoni301/Hackathon_1_TensorflowServing_WithoutDocker
VatsalSoni301/Hackathon_2_Scheduling
VatsalSoni301/HandWrittenDigitRecognition_Parallel_OpenMP
VatsalSoni301/MiniSQLEngine
VatsalSoni301/NeuralNetwork
VatsalSoni301/OverfittingAvoidTechnique
VatsalSoni301/Parallel_Assignment_1
VatsalSoni301/Parallel_Assignment_2
VatsalSoni301/pca_kmean_gmm
VatsalSoni301/PintOSScheduling
VatsalSoni301/PROBLEM-SOLVING
VatsalSoni301/ProtocolImplementation
VatsalSoni301/RNN_HMM
VatsalSoni301/RPC
VatsalSoni301/Server-Bootstrap
Server-Bootstrap
VatsalSoni301/VatsalSoni301.github.io