Using a custom Long short-term memory model in comparison with Meta's Prophet to predict the future price of Bitcoin with ablation study.
This assignment was completed as part of the coursework for my Bachelor of Science (B.Sc.) degree in Computer Science.
84/100 (First Class)
Feedback to Learner 15/12/23 16:09 The explanatory analysis and preprocessing is nicely performed and contains insightful visualizations. The formatting of the report could have been improved. The modeling part using LSTM and Prophet seems approprate. Nice ablation study comparing batch size against Prophet accuracy and number of units vs R2 score.
Dataset: 16/20
Modeling: 36/40
Code: 14/20
Report: 18/20