Sarah S. Lam

Breast cancer diagnosis based on feature extraction using a hybrid of K-means and support vector machine algorithms

With the development of clinical technologies, different tumor features have been collected for breast cancer diagnosis. Filtering all the pertinent feature information to support the clinical disease diagnosis is a challenging and time consuming task. The objective of this research is to diagnose breast cancer based on the extracted tumor features. Feature extraction and selection are critical to the quality of classifiers founded through data mining methods. To extract useful information and diagnose the tumor, a hybrid of K-means and support vector machine (K-SVM) algorithms is developed. The K-means algorithm is utilized to recognize the hidden patterns of the benign and malignant tumors separately. The membership of each tumor to these patterns is calculated and treated as a new feature in the training model. Then, a support vector machine (SVM) is used to obtain the new classifier to differentiate the incoming tumors. Based on 10-fold cross validation, the proposed methodology improves the accuracy to 97.38%, when tested on the Wisconsin Diagnostic Breast Cancer (WDBC) data set from the University of California - Irvine machine learning repository. Six abstract tumor features are extracted from the 32 original features for the training phase. The results not only illustrate the capability of the proposed approach on breast cancer diagnosis, but also shows time savings during the training phase. Physicians can also benefit from the mined abstract tumor features by better understanding the properties of different types of tumors.

Supply chain risk modelling and mitigation

In today’s global competitive environment, supply chains are more susceptible to vulnerability due to the increasing occurrence of internal and external risk events. In addition, the trend associated with lean management, which involves reducing inventory, leads to more dependency of supply chain partners on each other which exacerbates risk exposure of companies in the supply chain. This creates the need for more effective management of supply chain risks. In this research, a methodology based on Bow-Tie analysis and optimisation techniques is proposed to quantify and mitigate supply chain risks. The proposed methodology takes into consideration risk interconnections, and it identifies the best combination of mitigation strategies under budget constraints. A real case study from a high-end server manufacturing environment is presented. Results from the case study showed that the proposed methodology for risk modelling and mitigation can effectively be used to quantify the risks and achieve the required risk reduction at minimum cost while considering risk correlations.

Supply chain optimization under risk and uncertainty

We propose an approach for supply chain optimization under risk and uncertainty.Simulation and optimization models are combined through an iterative procedure.A software application is developed based on the proposed approach.To validate the proposed approach, a case study is provided. This paper presents an approach and a software application for supply chain optimization under risk and uncertainty. The proposed approach combines simulation and optimization techniques for managing risks in supply chains. A multi-objective optimization model is developed which considers the deterministic features of the supply chain. A simulation model is used to represent the stochastic features of the supply chain. Both models communicate to achieve the best values for profit, lead time and risk reduction by selecting a combination of mitigation strategies and allocating orders and inventory. A case study from a high-end server manufacturing environment is used to demonstrate the validity of the proposed approach. The analytical results show clear trade-offs among the three objectives where changing the risk reduction goal value will affect the total profit and lead time. The proposed approach helps decision makers identify the best risk mitigation strategies and allocate inventory and customer orders effectively.

Reliability and topology based network design using pattern mining guided genetic algorithm

Genetic algorithm guided with pattern mining for designing optimal network is proposed.Networks considered have different link and node failure rates.Association rules between networks of good and bad performances are devised.Mined patterns guide the genetic algorithm during mutation & crossover operations.Proposed method is effective in designing optimal network. This research proposes a new reliable network design methodology that is based on a pattern mining guided genetic algorithm (GA). The proposed method can be applied for a variety of applications including telecommunication, ad hoc, and power systems. In these networks, failures in certain parts of a network make it necessary for other parts to tolerate a higher traffic load in order to maintain adequate network connections. In addition, path changes due to dynamic routing of traffic can cause a time delay of communications in the network. To understand and reduce the connection failures costs, vigorous investigations are required to select the best design option under budget constraints. Given that many options for network topology and reliability allocation exist, a GA guided with pattern mining is proposed as an effective optimization method to design reliable network while considering link and node failures. Experimental designs under various assumptions have concluded that the guided GA approach is effective in identifying a network solution within a short period of time.

A decision support system for real-time order management in a heterogeneous production environment

We proposed a system for order management in heterogeneous production environments.The proposed system is supported by a case study form high-end server manufacturing.We presented an order prioritization tool to assess and prioritize customer orders.We used a risk mitigation approach to account for business risks.We developed a real-time dashboard to visualize order related information. In todays competitive market, many companies are morphing from the traditional new build, single brand, and silo environments to facilities accommodating diverse business missions. The later are called heterogeneous production environments in which the different business channels share their final production stage (shipping) to enable competitive advantages. In these production environments, at the operational level, the critical success factors are customer satisfaction, on-time delivery, product complexities, supply allocation, and resource utilization. At the strategic level, the success factors are revenue, customer urgency, and sales impact. This study proposes an End-to-End Customer Order Management System (E2E COMS) focusing on effective utilization of individual and shared resources to support real-time order management and mitigate risk of managing diverse missions. The proposed system consists of three integrated tools: Order Prioritization Tool (OPT) to assess and prioritize customer orders for each business channel, Order Fulfillment Progress Projection Tool (OFPPT) to predict the expected remaining order completion time considering inventory and resource capacity constraints, and risk mitigation tool to assess the risk of missing an order shipment due to shipping constraints. A real-time dashboard is developed to visualize the prioritized customer orders, expected time to arrive at the shipping area, shipping instructions, and two-dimensional risk assessment charts. The proposed system can effectively be used for shipping capacity management as well as prompt decision making.