Srinivasa Murthy Gunturu is a machine learning engineer with experience in exploratory data analysis, regression modeling, natural language processing, predictive analytics, and more. He has worked as a researcher developing NLP methods and as a data analyst developing classification and forecasting models. His education includes a Bachelor's degree in Computer Science and some published papers. Personal projects include predicting blood donations and weather forecasting using time series analysis and ARIMA modeling.
Srinivasa Murthy Gunturu is a machine learning engineer with experience in exploratory data analysis, regression modeling, natural language processing, predictive analytics, and more. He has worked as a researcher developing NLP methods and as a data analyst developing classification and forecasting models. His education includes a Bachelor's degree in Computer Science and some published papers. Personal projects include predicting blood donations and weather forecasting using time series analysis and ARIMA modeling.
Srinivasa Murthy Gunturu is a machine learning engineer with experience in exploratory data analysis, regression modeling, natural language processing, predictive analytics, and more. He has worked as a researcher developing NLP methods and as a data analyst developing classification and forecasting models. His education includes a Bachelor's degree in Computer Science and some published papers. Personal projects include predicting blood donations and weather forecasting using time series analysis and ARIMA modeling.
▪ Natural Language Processing ▪ BI Reporting ▪ Python & SQL ▪ Time Series ▪ Predictive Analytics ▪ Microsoft Azure
Education
Andhra University – Visakhapatnam, India April 2018
Bachelor of Technology – Computer Science and Engineering Career Chronology
Researcher | Department of Psychology – Texas Tech University | October 2018 – Present
▪ Developed machine-assisted methods used to analyze students’ written compositions utilizing NLP along with Prof. Roman Taraban, Ph.D ▪ Achieved seventy-two (72%) accuracy in classification, while human raters have shown seventy-five (76%) ▪ Published the paper in East European Journal of Psycholinguistics - http://eepl.at.ua/load/volume_6_number_1_2019/taraban_r_pittman_j_nalabandian_t_yang_w_f_z_mar cy_w_gunturu_s_m/210 ▪ Published a paper in American Society for Engineering Education (ASEE)- https://jee.org/35982 ▪ Applied LIWC, Naïve Bayes classifier, Text analysis Data Analyst | TATA Consultancy Services | June 2018 – Present ▪ Developed a Multi Class Ticket Classification Model on an Imbalanced Data set using XgBoost Algorithm and achieved Ninety- two (92%) Precession - Deployed the model using Kubernetes- Improved the response time by 80%-Servicenow api’s,NLTK,TF-IDF ▪ Developed a Chat-bot that address general queries, FAQ’s and also supports Incident Management using Azure Services and reduced the Incident count by 60%– Azure Bot Services, QnA Maker,Web- Apps,LUIS. ▪ Applied K-Means algorithm on “Server – logs” to logically group the similar scenarios - Elbow method is used to derive the optimal value of K - Clusters are visualized using PCA and t-SNE- Saved 40 hours per week for the backend/Support team. ▪ Analyzed the Impact of Weather parameters on products for a leading retailer - Secured the “Young Innovator Award” for the Europe Retail division – SQL, PySpark, Statistical Analysis ▪ Developed Item Matching model for a large ecommerce website along with a team of 3 members and achieved sixty-nine (69%) accuracy –Keras, Bidirectional -LSTM,NER,Encoder-Decoder
Personal Projects
Predicting Blood Donations | October 2017 – November 2017
▪ Participated in Driven Data competition in predicting the likelihood of any given person to donate blood - Applied Logistic Regression, Decision Trees - Listed in the top ten (10%) on the leader board with a log loss value of (0.445) Weather Forecasting Using ARIMA Model | December 2017 – April 2018 ▪ Collected the weather data from the Indian Meteorological Department (IMD) website and produced graphical images using heatmaps and ggplot – checked the stationarity via Dickey Fuller test and standardized time series by de-trending ▪ Plotted ACF / PACF charts to find all optimal parameters ▪ Analyzed time series data and forecasted temperatures for the coming year using the ARIMA model – the model with lowest AIC, BIC, and RMSE was chosen.