My-Data-Science-Projects

Welcome to my repo My-Data-Science-Projects. This repo contains the data science projects that I have worked so far. There are various topics in the projects including scientific computing, geotechnical research, civil engineering, health and finance. The computational methods incorporated in the projects are varied as well, including the following:

• Mathematical modelling.
• Numeric optimization.
• Machine learning implementations.

Most of the projects are machine learning implementations.

Table of Contents

1. Market Demand Analysis: Sales Forecast of a Coffee Shop Chain in the US

   In this project, we analyse a sales dataset from a coffee shop chain in the US, available in Kaggle's public datasets. We perform exploratory data analysis (EDA) and demand forecasting for the coffee shop, which is valuable for potential product portfolio optimisation. Several techniques from data analysis, statistics, and pure mathematics are leveraged in the process, highlighting the intricate interplay between abstract mathematics, computational methods, and market demand analysis. One of the key results of our analysis is that the demand for the products by the coffee shop chain is expected to increase by an average of 102.53% over the next six months. While the dataset provides valuable insights, it is important to note that its reliability may vary. Instead of focusing solely on the dataset's accuracy, this work demonstrates a unique blended multidisciplinary approach for market demand analysis with an emphasis on the methodology.

2. Forecasting PM2.5 Concentration on South Jakarta Air

   In this project, we perform a time series analysis on a dataset of Air Quality in South Jakarta imported from OpenAQ API. We analyse temporal patterns in the dataset and we perform a short period forecast using machine learning (ML) models as well as our package AbstrakTS.

3. Apple Stock Price Forecast

   In this project, we will perform a forecasting on Apple Inc. stock price for the 182 days ahead (6 months). The original dataset can be personally downloaded from Nasdaq. We will also demonstrate the implementation of a newly developed Python package for handling time series forecasting called AbstrakTS.

4. Concrete Strength Analysis

   In this project, we will analyse a concrete data set which consists of material compositions as features and the concrete strength data. First we analyse the relation of the features as well as the realation between each feature with the strength. Then we build a machine learning model to predict the strength of given concrete compositions.

5. Heart Disease Exploratory Data Analysis

   In this project, we perform an exploratory data analysis (EDA) on a dataset of heart disease. In the analysis, we uncover the most-likely factors correlated with heart disease.

6. House Price Prediction

   In this project, we develop a machine learning model for predicting house prices based on a dataset of house prices. We implement XGBRegressor algorithm from xgboost library for the ML model.

7. Nvidia Stock Price Forecast

   In this project, we perform a time series analysis on a dataset of Nvidia stock prices. We build a machine learning (ML) model for performing forecast on the stock price.

8. Geotechnical Research Projects

   In this directory, we present our research project in geotechnical engineering incorporating mathematical modelling, data analysis and machine learning implementations.