Stock Ranking Prediction Using List-Wise Approach and Node Embedding Technique

Traditional stock movement prediction tasks are formulated as either classification or regression task, and the relation between stocks not considered an input of prediction. The relative order ranking is more important than price return a single for making proper investment decisions. Stock performance can be improved by incorporating information in task. We employ graph-based approach use machine learning model. Investors might interested top- k they would profitable others. Thus, measure should top-weighted bounded any value k. Existing evaluation measures lack these properties, we propose new named normalized rank biased overlap ( NRBO@k) NRBO@k-based strategy generates 0.281% to 4.928% higher gain topmost stock-based strategy. show that list-wise loss function improve significantly approach. It better NRBO@10 combination point-wise pair-wise three out four cases. Node embedding techniques such Node2Vec reduce training time approaches significantly. Additionally, through hyperparameter tuning when sparse graph applied.