预答辩公告
论文题目 |
On Improving the Product Price Prediction Methods Using Deep Learning Approaches |
答辩人 |
Ahmed Fathalla |
指导教师 |
Professor KENLI LI |
答辩委员会 主席 |
|
学科专业 |
Computer Science and Technology |
学院 |
College of Computer Science and Electronic Engineering |
答辩地点 |
Online pre-defense |
答辩时间 |
2020年04月12日 |
学位论文简介
Due to today’s transition to online shopping in the last few decades, predicting product prices of e-commerce is gaining importance. Likewise, commodity prices are a significant factor for mining projects, as price volatility is an important parameter for mining company evaluation and investment decision making. In this work, we propose four contributions (i.e., prediction models) to address four price prediction problems using statistical, machine learning and deep learning methods. We utilized several algorithms from each domain.
The first two contributions address the price prediction problem of e-commerce products to fill in the gap of the current price prediction models for addressing second-hand items of e-commerce products.
The third and fourth contributions are introduced for addressing price forecasting of two vital energy markets, i.e., coal and oil markets using hybrid deep learning and ensemble learning methods, respectively.
In summary, the results demonstrate the superiority of the proposed methods over the baseline models. All the proposed methods lead to performance improvement.
主要学术成果
Ahmed Fathalla, Ahmad Salah, Kenli Li, Keqin Li and Piccialli Francesco, Deep end-to-end learning for price prediction of second-hand items, Knowledge and Information Systems, 2020. SCI IF:2.936 (Q1).
Zakaria Alameer, Ahmed Fathalla, Kenli Li, Haiwang Ye and Zhang Jianhua, Multistep-ahead forecasting of coal prices using a hybrid deep learning model, Resources Policy, 2020. SSCI IF:3.986 (Q1).
Ahmed Fathalla, Ken Li, Ahmad Salah and Marwa F Mohamed, An LSTM-based Distributed Scheme for Data Transmission Reduction of IoT Systems, Neurocomputing, 2021. SCI IF:4.438 (Q1).