Harisivamoorthi P.G

As a data enthusiast with two years of experience in SAP BO report development and production environment support, I excel in EDA, data visualization, and statistical analysis. My expertise in building and deploying machine learning models enables me to provide valuable insights to stakeholders and end-users.

Received Special Initiative Award :For Hardwork and effort behind Project for taken up the initiation on fixing the production issue, 2022.

Special Achievement Award: For providing support and Completing the task on time, 2022

Introduction to career skills in Data analytics

• Shinkansen Travel Experience - Hackathon Project

Objective: To predict the overall experience of the passengers based on attributes like 'Seat_Comfort',  'Leg_room', 'Onboard_service', 'Cleanliness' etc. 

Resolution: The classification algorithms that were used are Random Forest, CatBoostClassifier and  XGBoostClassifier. And finally, a comparison of accuracy across these models was done to finalize the model  for prediction. Achieved an accuracy of 94.68%.

Skills & Tools: EDA and Classification algorithms 

• Online Retail Orders Analysis using SQL

Objective: To utilize SQL skills to analyze an online retail store's order management database and provide  data-driven insights to help the company make informed decisions. 

Resolution: Successfully completed the project using SQL to analyze an online retail store's order management  database. Generated insights to help the company make informed decisions, contributing to the overall  growth of the business. Refer here. 

Skills & Tools: Joins, Sub Queries, SQL-clauses-statements-conditions, SQLite using DB Browser and MySQL  Workbench 

• Election Exit Poll Prediction and U.S.A Presidential Speech Analysis

Objective: The project involves two case studies in machine learning: Vote Prediction and Text Analysis. 

Resolution: The Vote Prediction task involves building a predictive model to determine a citizen's political  party based on age and survey responses, while the Text Analysis task involves analyzing U.S.A. presidential  speeches to gain insights. The results of the project are a predictive model for Vote Prediction and insights and  inferences drawn from the analysis of U.S.A. presidential speeches for Text Analysis. Refer here. 

Skills & Tools: Text Mining Analytics, K Nearest Neighbour - Naive Bayes, Ensemble Techniques, Logistic  Regression - Linear Discriminant Analysis 

• Predictive Modelling for Computer Usage and Women’s Contraceptive Usage

Objective: To develop predictive models using linear regression and classification techniques to make accurate  predictions for computer usage and women's contraceptive usage. 

Resolution: Successfully completed a project on predictive modelling using linear regression and classification  techniques to predict computer usage and women's contraceptive usage. Developed a linear regression model  to predict the percentage time the computer remains in user mode, achieving an accuracy of 85%.  Implemented classification techniques to classify whether women use or not use contraceptives, achieving an  accuracy of 78%. Utilized Python and scikit-learn libraries for data cleaning, pre-processing, and modelling.  Strengthened skills in predictive modelling, data analysis, and data visualization. Refer here. 

Skills & Tools: Linear Regression, Logistic Regression, CART, LDA 

• Segmentation using Clustering and PCA for Digital Marketing Advertisement and Primary Census Data

Objective: To perform data segmentation using clustering techniques and principal component analysis (PCA)  for digital marketing advertisement data and primary census data. 

Resolution: Successfully completed a data mining project on segmentation using clustering and PCA  techniques for digital marketing advertisement data and primary census data. Performed exploratory data  analysis (EDA) to understand the data and identify relevant variables. Used clustering techniques to segment  the digital marketing advertisement data, achieving an accuracy of 90%. Conducted PCA on the primary census  data to identify optimum principal components that explain the most variance in the data, achieving an  explained variance of 95%. Utilized Python and  scikit-learn libraries for data cleaning, pre-processing, and  modelling. Strengthened skills in data mining, clustering, PCA, and EDA. Refer here. Skills & Tools: EDA, Clustering, PCA, Data Mining, Silhouette Score, Segmentation 

scikit-learn libraries for data cleaning, pre-processing, and  modelling. Strengthened skills in data mining, clustering, PCA, and EDA. Refer here. Skills & Tools: EDA, Clustering, PCA, Data Mining, Silhouette Score, Segmentation