Data Analyst
Data Science aspirant based on DFW
A passionate individual for programming with a strong foundation in math, analytics and the ability to work autonomously having two years of experience in Application Development. I love problem solving and bringing the brains behind the solution.
Great passion and curiosity towards data analysis, big data and data science. I believe insights from big data can be a life-changer in the future. Exploring for an internship in the Fall 2020/Summer 2021 or a full time role after I graduate.
Designed linear regression model, implementing the gradient descent algorithm and experimented with different values of threshold and learning rate.
Compared performance with SVM kernels, Decision tree and found RBF kernel to have the highest accuracy of 98%.
Performed feature selection, used the transformed features on the neural network to achieve accuracy of 74.90%.
Analyzed daily observations from Australian weather stations with variables like windspeed, humidity, temperature etc and predicted if it will rain the next day.
Implemented SVM, Decision tree and boosting and achieved highest accuracy of 84.82% using Gradient boosting since it learns from weak classifiers and improvises each stage
Performed ANN using Tensor flow, Keras API and experimented with KNN to achieve 83.38% accuracy.
Analyzed the IMDB Reviews Dataset that contained the reviews and sentiments using nltk and sklearn packages.
Designed wordcloud to witness the frequent positive and negative reviews.
Preprocessed the dataset by tokenizing the words, removing stopwords, html strips and noisy texts, stemmed the words using Portable Stemmer.
Performed TFIDF score which penalizes the frequent words and achieved an accuracy of 75% to predict the number of positive and negative reviews by using Logistic Regression, Naive Bayes algorithm
FirstNet is operated to public safety allowing them to get information quickly during natural calamities and helping them making decisions.
Collaborated in designing automation techniques using Robot framework with Selenium for testing purposes which reduced the run processing time by 40% and minimized errors by 90%.
Implemented test scenarios using Safe Agile approach, identified the E2E scenario, created the test plan and presented the modules to client during deployment of each sprint.
Worked with Web and Mobile Cato tool to ensure the webpage accessible for physically challenged people.
Visualized different natural calamities on the map that affect the respective counties at United States.
Collected data on different Exxon refineries in the USA, located their counties and analyzed which all disasters are impacting the refinery locations.
Identified the top 4 disasters and presented it as a story. This helps to know the advent of disasters to refinery companies before hand and they could plan appropriate measures to tackle them.