Name: Anuja Dixit
Type: User
Company: University Of California
Bio: MS- Business Analytics | MBA - Finance | BE- E&TC | Work Exp- 7+| Jobs- BA (Amazon), Product Analyst (SEW), SDE (TCS) | Interests- Data, Product, Mgm
Location: Irvine, California
Blog: www.linkedin.com/in/anuja-dixit
Anuja Dixit's Projects
Analyzed the Twitter Reviews for the 5 major US Airlines and found insights on the improvements. Used Natural Language Processing in Python to classify an airline review text. Trained the model using a variety of machine learning algorithms such as logistic regression, random forest, knn for tweet classification prediction and tested using model evaluation techniques such as holdout evaluation, cross-validation, ROC and AUC curve to conclude the best model. Used grid search method for hyperparameter tuning to find the best tuned model with an accuracy of 77.4%
Analysed global superstore data and drew insights using exploratory data analysis
Analyzed a Portuguese bank dataset to determine whether a customer would agree to a term deposit or not in a time span of less than 24 hours using machine learning techniques of Random Forest and Logistic Regression classification in R.
Analyzed an HR dataset with the aim to help companies to solve the employee attrition problem. Data exploration analysis was conducted on the dataset to derive insights and a classification model was recommended to predict whether an employee is likely to leave the company.
Forecasting and Predicting MCD's stock price on December 31, 2020 using MCD's historical stock price and economic drivers from FRED and other sources. Building a what-if dashboard for the same.
Analyzing the impact of factors like price, age-rating, recent updates, languages on the ratings of the mobile game app.
Predicting the genre of a song based on its lyrics using Natural Language Processing
Analyzed the regular season and play off data for the NFL and predicted the scores for the final Superbowl 54. Ensemble model using Alteryx and R predicted Kansas City Chiefs to win with a score of 30. Voila, Kansas did win with a score of 31
Forecasted using ARIMA & ETS models how many units of a particular product will be purchased from retailers locations based on a historical trend.