Search results

This paper considers the complex stochastic systems such as supply chains, whose dynamics are controlled by an unknown parameter such as the arrival or service rates. The purpose of this paper is to provide a simulation-based estimator of the unknown parameter when only partially observed data on the underlying system is available.

The proposed method treats the unknown parameter as a random variable and estimates the parameter by computing the conditional expectation of the random variable given the partially observed data. This study then express the conditional expectation as a weighted sum of reverse conditional probabilities using Bayes’ rule. The reverse conditional probabilities are estimated using simulation.

The simulation studies indicate that the proposed estimator converges to the true value of the conditional expectation as the computer time allocated to the simulation increases. The proposed estimator is computed within a few seconds in all of the numerical examples, which demonstrates its time efficiency.

Most of the existing methods for estimating an unknown parameter require a significant amount of simulation, causing long computation delays. The proposed method requires a single simulation run for each candidate of the unknown parameter. Thus, it is designed to carry a significantly reduced computational burden. This feature will enable managers to use the proposed method when making real-time decisions.

The purpose of this paper is to identify the root cause of low yield problems in the semiconductor manufacturing process using sensor data continuously collected from manufacturing equipment and describe the process environment in the equipment.

This paper proposes a sensor data mining process based on the sequential modeling of random forests for low yield diagnosis. The process consists of sequential steps: problem definition, data preparation, excursion time and critical sensor identification, data visualization and root cause identification.

A case study is conducted using real-world data collected from a semiconductor manufacturer in South Korea to demonstrate the effectiveness of the diagnosis process. The proposed model successfully identified the excursion time and critical sensors previously identified by domain engineers using costly manual examination.

The proposed procedure helps domain engineers narrow down the excursion time and critical sensors from the massive sensor data. The procedure's outcome is highly interpretable, informative and easy to visualize.