Binary Classification for Next-Day Gold Price Movements Using Machine Learning

Next-Day Gold Price Movements

My contributions began at the stage of formulating the research idea and designing the methodology. I was involved in the brainstorming process to determine the scope of the research, the selection of a classification approach over price regression, and the design of the feature engineering used to represent gold price dynamics.

On the technical side, I contributed to the data preparation process, feature selection, machine learning code development, model experimentation, and result visualization. I was also involved in the selection and evaluation of the machine learning models used in the research, namely Naïve Bayes, Logistic Regression, and Support Vector Machine. Additionally, I performed hyperparameter tuning for all three models to obtain optimal model configurations and analyzed the impact of parameters on classification performance.

In the manuscript preparation, I was responsible for writing the Abstract section, a portion of the Literature Review discussing the machine learning models used, and the Machine Learning Methods section. I also created various graphs and visualizations used to explain the results of exploratory data analysis, feature engineering, and model evaluation results presented in the paper.