gaurav

Data scientist
placeholder
Location Banglore
view-icon
Total Views196
availability
Shortlist0
back in time
Member since30+ Days ago
 back in time
Contact Details
phone call {{contact.cdata.phone}}
phone call {{contact.cdata.email}}
Candidate Information
  • User Experience
    Experience 6.5 Year
  • Cost
    Hourly Rate$10
  • availability
    AvailabilityImmediate
  • work from
    Work FromAny
  • check list
    CategoryInformation Technology & Services
  • back in time
    Last Active OnSeptember 25, 2022
Key Skills
Data sciencePythonHTMLJavaScript
Education
2014-2019

(Data science)
Excelr solutions

Summary
I am passionate towards developing simple, transferable ML models that provide time bound technological insights for business and love building exemplary data science teams to achieve greater goals. • Implemented Analytics solutions for Banking, E-commerce and real estate domain. • Significant experience with Scikit-Learn, nltk, Scipy, Numpy, Pandas, Exploratory Data Analysis • Proficiency in Flask, Heroku and HTML etc for deploying machine learning models at scale. • Interested in creating a distributed system for training machine learning algorithms. • NLP Analytics: Text Mining, Sentiment Analysis, Bag of words modelling.
Project Details
Title :project is to predict whether the customer will fall under loan default or not
Duration :1
role and responsibileties :

data scientist

Description :

·      We got the data set having 150000 observations.

·      The project life cycle includes the following steps: Data processing, cleaning, EDA, Model building, deployment.

·      The programming language used is python and we uploaded the data from CSV.

·      The data cleaning consists of dropping of NA values and resetting of index.

·      In EDA part we compared the relationship between our outcome variable with independent variables using various plots and checked the collinearity among the independent variables. 

·      In model building we tried the models: Logistic Regression, Decision Tree, Random Forest and XGB and we got the highest accuracy with XB boost with train accuracy: 95% and test accuracy: 93%.

We deployed the model using python flask and Heroku


Copyright : 2022 – OnBenchMark All Right Reserved.