Data Scientist

Technical Skills: Tensorflow, Python, R, SQL, Tableau, Matlab

Projects

Concentrate Cake Classification using CNN

Developed a model to classify concentrate cake quality based its video using tensorflow. Reached training accuracy 91% and validation accuracy 82%. The data is confidential.

Pipe Slurry Thickness Prediction using Support Vector Regression

Developed a model to predict the pipeline thickness using Support Vector Regression and reached 4% MAPE. The model is used to find best parameter so that tha slurry thickness reduction is minimum using Solver Excel.

Tobacco Garangan Classification Model using MobileNet CNN

Developed a mobile App that can classify Tobacco Garangan quality based its images using tensorflow with team. We found that the best pre-trained model to classify the case is MobileNet. We reached training accuracy 98% and validation accuracy 97%.

Demonstration Video

Water Feasibility Classification

Using Python and neural random forest to classify water feasibility (binary) with 38 factors consisting of 11990 entries. Reached 90.83% training accuracy.

Predicting Value at Risk with GUI Matlab

Used Matlab to make Graphical User Interface (GUI) that can predict Value at Risk (VaR) with 4 different methods.

K-Means Clustering with GUI R

Used R to make Graphical User Interface (GUI) that can cluster data with K-Means Clustering method.