Candidates About

Nirmal Manavalan

PROFESSIONAL SUMMARY

·            Over 1+ year of experience in data analysis and machine learning techniques with hands-on-experience on various python libraries like Numpy/Pandas/SciPy

·            Over 1+ years of hands-on-experience on python web application development using Django framework and front-end technologies like HTML/basic JavaScript, CSS for dynamic UI design.

·            Large data sets manipulation, data cleaning, quality control, and data management (Pandas)

·            Experience with Statistical analysis, Stat packages mainly in python

·            Hands-on-experience with visualization libraries in python like Matplotlib and Seaborn for designing graphs for data sets according to the business needs

·            Used tableau to analyze and obtain insights into large datasets, create visually compelling and actionable interactive reports and dashboards

·            Experience in writing Sub Queries, Stored Procedures, Triggers, Cursors, and Functions on MySQL database

·            Hands-on-experience with pandas, SciPy and Numpy packages in Python for data analytics

ACADEMIC BACKGROUND

·            SAN DIEGO STATE UNIVERSITY     MASTER OF SCIENCE [COMPUTER NETWORK] (May 2017)

·            ANNA UNIVERSITY                         BACHELOR OF ENGINEERING [ELECTRICAL ENGINEERING] (Aug 2014)

TECHNICAL SKILLSET

Programming Languages                   Python, C++, Java

Python Data Analytics libraries         Numpy/Pandas/SciPy

Visualization libraries in Python       Matplotlib/Seaborn/Plotly

Web Framework                                 Django, Flask

Python IDE                                           Sublime Text

Database                                             MySQL/SQLite

PROFESSIONAL EXPERIENCE

GIFT INFO TECH LTD. (CHENNAI, INDIA)                                                                                        Aug'2013 to May'2014

Role: Python developer

·   Developed a prototype website for a shopping cart for the client

·   For designing website, Django framework was used with the MVT (Model View Template) architecture

·   Designed the front-end views with DHTML and Java script along with CSS Bootstrap, used MySQL for the backend database and wrote queries and stored procedures to deliver the data to the front end according to the client requirements.

·   Used Django Templating language for coding the HTML pages and used Django model forms in the project

INTEGER CORP (Buffalo, NY)                                                                                                                Jun’2017 to Nov’2017

Role: Python Data Analyst (Predictive Modeling)

●      Analyzed large data sets using pandas and used regression models using SciPy to predict future data and visualized them

●      Collaborated with data engineers, wrote and optimized SQL queries to perform data extraction from SQL tables

●      Analyzed large data sets to find any pattern in the data by extracting the data, cleaning the outliers and plotting the data using Seaborn library

●      Wrangled data, worked on large datasets (acquired data and cleaned the data), analyzed trends by making visualizations using matplotlib using Python

●      Used Linear Regression for predicting the data and Logistic Regression for classifying the predictor variable according to the classes.

RELEVANT COURSES

Python for data Science and Machine learning Bootcamp – UDEMY

RELEVANT PROJECTS

TITANIC DATASET – KAGGLE

Information in the dataset [Index, Passenger ID, Survived or not, Passenger Class, Name, Sex, Age, Ticket, Fare, Cabin, Embarked]

Basic Idea is to predict whether the person in the ship has survived or not using the logistic regression algorithm

·         Worked with the semi cleaned version of the titanic dataset to assess the people that survived

Exploratory Data Analysis

·         Exploratory data analysis was done to see the missing data. For this purpose, heat map plot was used to see the data missing in the entire dataset and where the data cleaning is to be made

·         Using seaborn library in Python the Count plot was used to assess the survival rate against the Sex(Male/Female) and the passenger class

Data Cleaning

·         From the plots there were two columns missing data, one column was cleaned by using values from adjacent rows, while the other column was dropped as it cannot be recovered.

·         A custom function was written to fill the missing Age values. After this the function was applied to the entire dataset and a heatmap plot was used to see if the data set has any missing values

·         After this the categorical values were converted so that the data will fit to the training model

Converting Categorical Features

·         Converting the categorical features to dummy variables using pandas so that the ML algorithm will be able to take the features as inputs

Building a Logistic Regression Model

·         The data is split into training and testing set. The training set is fit into the Logistic regression model

·         The predictor variable is also assigned, and the model is now ready to be evaluated

Evaluation

·         Precision is evaluated using the metrics. The confusion matrix can be directly retrieved, or the classification report is printed using the test model and predictor variable