How can we accurately predict the performance of a Personal Computer (PC) configuration without time consuming simulation? In this work, we predict the performance of a computer hardware configuration using Multiple Neural Networks (MNN). We use Principal Component Analysis (PCA) during data preprocessing as guidance for model creation. The input data includes the internal component characteristics of a computer. A deep learning model is used to infer a benchmark score given a hardware configuration. The finished model takes input data such as Central Processing Unit (CPU) type, frequency, number of cores, memory size and speed, flash or disk architecture, network configuration and correlates it against the corresponding performance benchmark value and system response to a benchmark workload. We demonstrate the accuracy and effectiveness of the MNN and PCA machine learning models using the Standard Performance Evaluation Corporation (SPEC) benchmarks (SPEC CPU2006 and SPEC CPU2017), and a set of approximately 50,000 commercial machines configurations. Our MNN model is able to achieve an average accuracy rate of 97.5% for all benchmarks. Our results provide both personal and enterprise users a tool that can accurately estimate system configuration performance without lengthy and resource intensive benchmarking sessions.

Extract, Transform and Load (ETL) is a fundamental process used to populate a data warehouse and an important step towards data integration which is a key step for data preprocessing for many data mining projects. The objective of this project is to create a simple web based ETL tool which can be accessed by anyone with internet access. This ETL tool is also supported by a brief on-line tutorial on ETL process. The combination of tool and tutorial makes an ideal self-paced interactive learning tool for a beginning ETL learner. Guests will have limited access to this tool and registered users can have complete access. Using this tool, data can be extracted from text files or MySQL tables, integrated and loaded into target MySQL tables. Before loading the data into target MySQL tables, various transformations can also be applied to them.

Vital statistics data offer a fertile ground for data mining. The authors discuss the results of a data mining project on the causes of death aspect of the vital statistics data in the state of California. A data mining tool called Cubist is used to build predictive models out of two million cases over a nine-year period. The objective of our study is to discover knowledge that can be used to gain insight into various aspects of mortality in California, to predict health issues related to the causes of death, to offer an aid to decision- or policy-making processes, and to provide useful information services to the customers. The results obtained in our study contain valuable new information.

This paper reports our experimental results in protein secondary structure prediction using the machine learning software, C5. The accuracy improvement in the prediction of protein secondary structure is the focus of our study. Starting with a target protein with unknown secondary structures, we investigate three different approaches and find that training cases selected based on sequence homology can achieve the highest predictive accuracy of 75% in testing cases. Our result indicates that the method of selecting proteins for the training cases has the most significant impact on predictive accuracy.

A model for analyzing a FCS (fault-tolerant clock synchronization) system of the type supported by a statistical self-diagnosis is described. Once a self-diagnosis scheme is integrated into an FCS design, the problem of controlling and measuring the system's self-stability arises. A stochastic Petri net (SPN) model is constructed to derive the self-stability measures of such FCS systems. An example is given to demonstrate the entire modeling and analyzing procedure. The mapping from SPN model to Markov model shown in an example can be automated by using an SPN software package. The results show that the SPN model is an excellent tool for obtaining self-stability measures and that several important system features, such as synchronization and parallelism, can be modeled using the SPN method in a much clearer manner than they can be modeled using other available tools.>