Sourabh

Dynamic and results-oriented professional with extensive experience in data analysis, software development, and database management. Proven ability to collect, clean, and interpret complex data sets, alongside a strong foundation in requirement analysis. Self-directed and proactive, I excel in leveraging cross-functional knowledge of information technology and business processes to drive impactful solutions.

FOOD VOLUME ESTIMATION :

Business Problem: To estimate the volume of food by an image.

Data Collection: Google,Manually,and rest of the data from Food101 kitchen dataset. 

Model fitting: MaskRCNN,CNN,Resnet101,Monocular Depth Estimator .

Model performance evaluation: Evaluating model performance by Accuracy.

Tools Used : python,pandas,Sklearn,Tensorflow

WINE QUALITY PREDICTION :

Data Pre-Processing: Use of statistical value for data imputation to replace missing values basis quantum  of such observation, variable distribution and variable type .

Exploratory Data Analysis: Performed univariate and bivariate analysis .

Model fitting : Linear regression model and Decision tree model .

Model performance evaluation: Model performance basis MAPE, R square.

Tools Used : python,pandas,Sklearn,Tensorflow.

DIABETES PREDICTION :

Exploratory Data Analysis(EDA): Performed univariate, bivariate and multivariate analysis 

Model fitting: Decision Tree -Regressor, performed hyperparameter tuning,KNN model 

Model performance evaluation: model performance and performed hyperparameter tunning

Tools Used: python,pandas,matplotlib.pyplot,sklearn,seaborn

EDA on Spotify dataset :

Created insights on top energetic songs,loudness.

Understood the relationship between different features. 

Handles null values and outliers.