Arifsofyan (Arifsofyan)

2023: In this project I utilised various feature engineering techniques such as feature modification, data transformation, and missing value handling to build a dataset ready for the modelling process.

Through careful feature selection and predictive modelling using the Random Forest Classifier algorithm, I managed to build a model that can predict loan eligibility with high accuracy. The next step was to conduct a thorough evaluation of the model, including performance testing on test data to ensure its reliability and validity.

This project not only broadened my understanding of data analysis and predictive modelling, but also enriched my practical experience in handling large and complex datasets and honed my feature engineering and modelling skills. The success of this project demonstrates my contribution in supporting smarter and more effective decision-making in finance.

2024: In this project, I conducted in-depth data analysis to identify the factors affecting dropout rates in educational institutions. Then created an interactive dashboard to effectively visualise the data findings. In addition, developing a reliable machine learning model to forecast the likelihood of students experiencing dropout, by ensuring the model does not experience overfitting or underfitting. This project combines data analysis, machine learning model development, and practical solution implementation to improve the efficiency and sustainability of educational institutions.

2024: This project aims to address the challenge of high employee turnover rates that impact recruitment costs, team stability, and operational efficiency within the company. Through in-depth data analysis and the implementation of cutting-edge technologies such as machine learning and interactive dashboards, we seek to identify key factors influencing attrition and provide strategic recommendations to retain valuable workforce and enhance productivity.

2024: This project aims to develop an image classification model using deep learning techniques to recognise different types of food. The model is built using the Food11 dataset which consists of over 16,000 food images divided into 11 different categories. And managed to get a model with 93% accuracy