"Truth is what stands the test of experience."    - Albert Einstein

 Data Science in Manufacturing (製造數據科學)/ Intelligent Manufacturing Systems (智慧型製造系統)

Manufacturing Data Science (MDS) is a decision-oriented system which has the computational intelligence and self-learning ability to optimize the manufacturing process. The techniques include data science (decision tree, deep learning, feature engineering, clustering, etc.), meta-heuristic algorithms (e.g. tabu search, simulated annealing, genetic algorithm, particle swarm optimization, etc.), fault detection & classification (FDC), statistical process control (SPC), engineering data analysis (EDA), etc. These methodologies can optimize resource allocation and support trouble-shooting process. The applications are diversified such as capacity planning, production scheduling, machine configuration optimization, process fault diagnosis, facility layout, bottleneck identification, quality and yield enhancement, market segmentation, pattern recognition, etc. Real-setting empirical studies were conducted to validate the proposed model and improve the business core competence effectively.

Research Interests:
  • Data Science in Manufacturing (製造數據科學)
    • Data Mining for Yield Improvement (數據探勘與良率改善)
    • Statistical Process Control (SPC) Big Data Analytics (統計製程管制大數據分析)
    • Process Diagnosis & Pattern Recognition (製程診斷與樣型識別)
    • Price Forecasting and Optimal Decision by Reinforcement Learning (價格預測與強化學習)
    • Virtual Material Quality Investigation (虛擬物料品質檢測)
    • MECE Engineering Feature Selection and Predictive Maintenance (工程參數篩選與預測保養)

  • Manufacturing System Management (製造系統管理)
    • Multi-Objective Job-Shop Stochastic Scheduling (多目標隨機生產排程)
    • Demand Forecasting and Robust Capacity Planning (需求預測與穩健產能規劃)
    • Work Study and Time Motion Study (工作研究與時間動作研究)
    • Vendor Selection and Order Allocation in Supply Chain (供應鏈廠商評選與訂單配置)
    Representative Publications:

      Productivity and Efficiency Analysis (生產力與效率分析)/ Production Economics (生產經濟學)

    Productivity and efficiency analysis (PEA) is a discipline to assess the performance of production system and drive productivity. The technique is developed to estimate production function based on production economics and then the efficiency of production system can be identified. PEA supports evaluating the productive efficiency, productive effectiveness, product mix, scale size, productivity change, performance benchmarking, market power, etc. The research focuses on developing nonparametric method (e.g. Data Envelopment Analysis, DEA) and semi-parametric method (e.g. Stochastic semi-nonparametric envelopment of data, StoNED) applied to the areas of manufacturing, airlines, energy market, power system, cap-and-trade, carbon emission, biofuel diesel, etc.

    Research Interests:
    • Price for CO2? 一噸二氧化碳賣多少錢?
    • Data Envelopment Analysis (數據包絡分析)/ Performance Evaluation (績效評估)
    • Stochastic semi-nonparametric envelopment of data (StoNED)
    • Marginal Abatement Cost and Allocation of Emission Permit (邊際減排成本與排放權配置)
    • Productive Efficiency, Effectiveness and Scale (生產效率、有效生產與最適規模)
    • Marginal Profit/Productivity Analysis (邊際利潤/邊際生產力分析)
    • Nash Equilibrium in Oligopolistic Energy Market (能源寡占市場下之納許均衡)
    Representative Publications: