Machine Learning for Earnings Prediction: A Nonlinear Tensor Approach for Data Integration and Completion

Computer Science

• Business Problem 100%
• Convolutional Neural Network 100%
• Data Integration 100%
• Firm Characteristic 100%
• Industry Sector 100%
• Information Loss 100%
• Machine Learning 100%
• Performance Metric 100%
• Portfolio Return 100%
• Prediction Accuracy 100%
• Prediction Error 100%
• Predictive Model 100%
• Quality Dimension 100%

Keyphrases

• Accuracy Improvement 20%
• Analyst Forecast Accuracy 40%
• Business Problems 20%
• Consensus Forecasts 40%
• Convolutional Neural Network 20%
• Data Completion 100%
• Data Integration 100%
• Data Quality 40%
• Earnings Prediction 100%
• Financial Analysts 40%
• Financial Applications 20%
• Firm Characteristics 20%
• Forecast Data 20%
• Heterogeneous Data 20%
• High Heterogeneity 20%
• Holding Period 20%
• Imputed Data 20%
• Industry Sector 20%
• Long-Short 20%
• Machine Learning 100%
• Multiple Datasets 20%
• Neural Network 20%
• Nonlinear Tensor 100%
• Performance Metrics 20%
• Portfolio Returns 20%
• Prediction Accuracy 20%
• Prediction Error 20%
• Predictive Models 20%
• Redundancy 20%
• S&P 500 20%
• Single Model 20%
• Stock Market Information 20%
• Structure Dimension 20%
• Tensor Approach 100%

Economics, Econometrics and Finance

• Earnings Announcement 66%
• Industry 33%
• Information Market 33%
• Machine Learning 100%
• Profit 100%