![]() □ Exciting News! Boosting Model Accuracy: The Role of Hyperparameter Optimization in Machine Learning is now live! □ The paper is available from the link posted in the first comment.īoosting Model Accuracy: The Role of Hyperparameter Optimization in Machine Learning Their empirical results showed that the stock selection method based on CNN outperformed other baseline methods (decision tree, support vector machine, and feedforward neural network) in terms of the annual return, sharpe ratio, and max drawdown. The pseudocodes of their procedures are listed on pages 4 (Algorithm 1) and 6 (Algorithm 2) that interested developers can implement which is straightforward. Second, they constructed a “factor picture,” and then took advantage of CNN’s image classification capability to achieve the purpose of stock selection. The feature extraction of stock selection factors is realized and then finally constructed a multifactor stock selection data set. In their workflow, they first calculated the correlation between stock selection factors and then combined the factors with low correlation. classified them by CNN (convolutional neural network) to select stocks.built a stock multifactor data set, and then constructed a “factor picture”.To address the two main problems mentioned above, the authors of : The traditional ML methods can be difficult in extracting features, which can lead to poor classification.The feature extraction of high dimensional financial data is still a complex work, which makes it difficult to select influential factors (out of many).Most stock selection methods available today have achieved good experimental results, but there are a couple of problems that exists: This means that a wrong transaction may cause large losses in which most of this loss is irreversible. Ordinary investors have a low fault tolerance rate. ![]() Financial investment is an area with relatively small fault tolerance. ![]() These methods achieve different metrics performance like annual return, sharp ratio, max drawdown, etc. There are tons of different methods that have been proposed over the years to improve this task, ranging from Statistical, Econometrics and Machine-/Deep-Learning (ML/DL) models, especially in the last decade or so. ![]() This task of stock selection is a very challenging one. Therefore, the selection of shares is subject to fulfilling a number of criteria. The performance of a stock depends on a number of criteria based on the risk-return measures. Stock selection's main objective is to distribute capital to selected stocks to get the most profitable returns at a lower risk. Stock ranking/selection for portfolio construction is an effective method for screening high investment value stocks in the future and can strongly assist investors in making decisions. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |