Certifications/Projects
Simple Counter Design using React JS
Created a simple Counter App using React JS. Understood the concepts of Virtual DOM and Physical DOM and the communication between them. Used Visual studio code for the project. Got hands on experience on React JS by developing using components using classes and also Stateless Functional Components. Understood React JS well as part of this project.
Keywords: React JS
Project Timeline: Feb 2020
Technologies: React Js
IDE: Visual Studio
Computation of the Capacity of National Aerospace system - Algorithms
Computed the capacity(maximum number of passengers travelled) between Los Angeles and Newyork city. There are few intermediate airports considered during the process and the time constraint for this problem is one day i.e Jan 6 2020. A Time based graph has been generated with 24 hours in the day as 24 nodes between the cities and at each hour/node the intermediate cities are considered. Then Ford Fulkerson algorithm(Network Flow Algorithm) is applied to compute the Maximum capacity and cut theory is applied to verify the same. Python NetworkX package is used during the process for generating the nodes and computing the max_capacity.
Keywords: Python 3.7
Project Timeline: October 2019 - November 2019
Technologies: Python
IDE: Spyder
Prediction of Heart disease using various Machine Learning classifier Algorithms
Problem Description : In today’s modern world, cardiovascular disease is currently one of the leading cause of death across the globe. Diagnosing the patients correctly on timely basis is the most challenging task for medical fraternity. Thus, it’s an implicit necessity to predict the condition at the earliest. The objective of this project is to build classifiers to predict the possibility of predicting whether the person is getting cardiovascular disease Supervised Learning Algorithms used : 1) Decision Tree (CART - Classification and Regression Trees) 2) Random Forest(Ensemble learning with multitude Decision trees - voting concept) 3) K-nearest neighbor algorithm ( Chose hyper parameter as K= 7 and metric/distance = minkowski) for getting higher accuracy. 4) Naive Bayes.
Dataset : Cleveland Clinical Foundation UCI Machine Learning Repository: http://archive.ics.uci.edu/ml/datasets/heart+disease Results: Got accuracy around 89% for Decision trees and similar amount for other algorithms as well.
Experiment : Tested on epilepsy data set and got accuracies around >95%.
https://github.com/hackin123/hackin123.github.io/blob/master/files/Prediction%20of%20Heart%20and%20Epilepsy%20diseases%20using%20Machine%20Learning%20Classifiers.pdf
Keywords: Python 3.7
Project Timeline: October 2019 - November 2019
Technologies: Python
IDE: Spyder
Divide and Conquer, Sorting and Searching, and Randomized Algorithms
The primary topics in this part of the specialization are: asymptotic (“Big-oh”) notation, sorting and searching, divide and conquer (master method, integer and matrix multiplication, closest pair), and randomized algorithms (QuickSort, contraction algorithm for min cuts).
Keywords: Python 3.7
Project Timeline: June 2019 - August 2019
Technologies: C++,Python
IDE: PyCharm, Jupyter Notebook,Dev cpp
Introduction to Time Frequency Analysis AND Wavelet Transforms
Time-frequency analysis (TFA) is concerned with simultaneous analysis of signals in time-and frequency-domains. A widely encountered topic in signal processing, TFA tools are the standard for the analysis of multiscale systems (e.g., speech processing, seismic data analysis, multiscale filtering, etc.). In this course, I learnt about the basic concepts of time-frequency analysis and three widely established methods, namely, the short-time Fourier transform (STFT), Wigner-Ville distributions and wavelet transforms, starting with a review of Fourier transforms. Emphasis was laid on wavelet transforms and their applications. Theoretical concepts are learnt with practical demonstrations in MATLAB at each stage of the course. I used the concepts learnt in this course while publishing the two papers. I have secured place in top 5 performers list in this course.
Keywords: Wavelet Transforms
Project Timeline: Jan - March 2015
Technologies: Matlab
Tortuosity measurement in retinal images for early detection of diabetic retinopathy and hypertension
As part of feature extraction and classification set of retinal images of patients are cosnidered and the tortuosity measurement of nerves is done to detect diabetic retinopathy and hypertension at early stages. Matlab image toolbox is used for this project.
Technologies: Matlab