scholarly journals Equipment Anomaly Detection for Semiconductor Manufacturing by Exploiting Unsupervised Learning from Sensory Data

Sensors ◽  
2020 ◽  
Vol 20 (19) ◽  
pp. 5650
Author(s):  
Chieh-Yu Chen ◽  
Shi-Chung Chang ◽  
Da-Yin Liao
Keyword(s):  
Anomaly Detection ◽  
Unsupervised Learning ◽  
Operations Management ◽  
Hypothesis Test ◽  
Scale Up ◽  
Sensory Data ◽  
Stacked Autoencoders ◽  
Sample Data ◽  
The Difference ◽  
Control Wafer

In-line anomaly detection (AD) not only identifies the needs for semiconductor equipment maintenance but also indicates potential line yield problems. Prompt AD based on available equipment sensory data (ESD) facilitates proactive yield and operations management. However, ESD items are highly diversified and drastically scale up along with the increased use of sensors. Even veteran engineers lack knowledge about ESD items for automated AD. This paper presents a novel Spectral and Time Autoencoder Learning for Anomaly Detection (STALAD) framework. The design consists of four innovations: (1) identification of cycle series and spectral transformation (CSST) from ESD, (2) unsupervised learning from CSST of ESD by exploiting Stacked AutoEncoders, (3) hypothesis test for AD based on the difference between the learned normal data and the tested sample data, (4) dynamic procedure control enabling periodic and parallel learning and testing. Applications to ESD of an HDP-CVD tool demonstrate that STALAD learns normality without engineers’ prior knowledge, is tolerant to some abnormal data in training input, performs correct AD, and is efficient and adaptive for fab applications. Complementary to the current practice of using control wafer monitoring for AD, STALAD may facilitate early detection of equipment anomaly and assessment of impacts to process quality.

Monotone Split and Conquer for Anomaly Detection in IoT Sensory Data

2021 ◽  
pp. 1-1
Author(s):  
Thien-Binh Dang ◽  
Duc-Tai Le ◽  
Tien-Dung Nguyen ◽  
Moonseong Kim ◽  
Hyunseung Choo
Keyword(s):  
Anomaly Detection ◽  
Sensory Data

Investigation of Regression-Based Effect Size Methods Developed in Single-Subject Studies

2021 ◽  
pp. 014544552110540
Author(s):  
Nihal Sen
Keyword(s):  
Effect Size ◽  
Experimental Studies ◽  
Single Subject ◽  
Single Subject Design ◽  
Data Set ◽  
Design Studies ◽  
Sample Data ◽  
The Difference ◽  
D Values ◽  
Effect Size Calculation

The purpose of this study is to provide a brief introduction to effect size calculation in single-subject design studies, including a description of nonparametric and regression-based effect sizes. We then focus the rest of the tutorial on common regression-based methods used to calculate effect size in single-subject experimental studies. We start by first describing the difference between five regression-based methods (Gorsuch, White et al., Center et al., Allison and Gorman, Huitema and McKean). This is followed by an example using the five regression-based effect size methods and a demonstration how these methods can be applied using a sample data set. In this way, the question of how the values obtained from different effect size methods differ was answered. The specific regression models used in these five regression-based methods and how these models can be obtained from the SPSS program were shown. R2 values obtained from these five methods were converted to Cohen’s d value and compared in this study. The d values obtained from the same data set were estimated as 0.003, 0.357, 2.180, 3.470, and 2.108 for the Allison and Gorman, Gorsuch, White et al., Center et al., as well as for Huitema and McKean methods, respectively. A brief description of selected statistical programs available to conduct regression-based methods was given.

scholarly journals Image Anomaly Detection Using Normal Data Only by Latent Space Resampling

2020 ◽  
Vol 10 (23) ◽  
pp. 8660
Author(s):  
Lu Wang ◽  
Dongkai Zhang ◽  
Jiahao Guo ◽  
Yuexing Han
Keyword(s):  
Anomaly Detection ◽  
Normal Distribution ◽  
Autoregressive Model ◽  
Probability Model ◽  
Latent Space ◽  
Industrial Inspection ◽  
Detection Stage ◽  
Novel Method ◽  
The Difference ◽  
Reconstruction Model

Detecting image anomalies automatically in industrial scenarios can improve economic efficiency, but the scarcity of anomalous samples increases the challenge of the task. Recently, autoencoder has been widely used in image anomaly detection without using anomalous images during training. However, it is hard to determine the proper dimensionality of the latent space, and it often leads to unwanted reconstructions of the anomalous parts. To solve this problem, we propose a novel method based on the autoencoder. In this method, the latent space of the autoencoder is estimated using a discrete probability model. With the estimated probability model, the anomalous components in the latent space can be well excluded and undesirable reconstruction of the anomalous parts can be avoided. Specifically, we first adopt VQ-VAE as the reconstruction model to get a discrete latent space of normal samples. Then, PixelSail, a deep autoregressive model, is used to estimate the probability model of the discrete latent space. In the detection stage, the autoregressive model will determine the parts that deviate from the normal distribution in the input latent space. Then, the deviation code will be resampled from the normal distribution and decoded to yield a restored image, which is closest to the anomaly input. The anomaly is then detected by comparing the difference between the restored image and the anomaly image. Our proposed method is evaluated on the high-resolution industrial inspection image datasets MVTec AD which consist of 15 categories. The results show that the AUROC of the model improves by 15% over autoencoder and also yields competitive performance compared with state-of-the-art methods.

Ultrasonic Attenuation Based Inspection Method for Scale-up Production of A206–Al2O3 Metal Matrix Nanocomposites

2015 ◽  
Vol 137 (1) ◽  
Author(s):  
Jianguo Wu ◽  
Shiyu Zhou ◽  
Xiaochun Li
Keyword(s):  
Metal Matrix ◽  
High Performance ◽  
Ultrasonic Attenuation ◽  
Hypothesis Test ◽  
Scale Up ◽  
Al2o3 Nanoparticles ◽  
Quality Inspection ◽  
Ultrasonic Dispersion ◽  
Inspection Method ◽  
Metal Matrix Nanocomposites

A206–Al2O3 metal matrix nanocomposite (MMNC) is a promising high performance material with potential applications in various industries, such as automotive, aerospace, and defense. Al2O3 nanoparticles dispersed into molten Al using ultrasonic cavitation technique can enhance the nucleation of primary Al phase to reduce its grain size and modify the secondary intermetallic phases. To enable a scale-up production, an effective yet easy-to-implement quality inspection technique is needed to effectively evaluate the resultant microstructure of the MMNCs. At present the standard inspection technique is based on the microscopic images, which are costly and time-consuming to obtain. This paper investigates the relationship between the ultrasonic attenuation and the microstructures of pure A206 and Al2O3 reinforced MMNCs with/without ultrasonic dispersion. A hypothesis test based on an estimated attenuation variance was developed and it could accurately differentiate poor samples from good ones. This study provides useful guidelines to establish a new quality inspection technique for A206–Al2O3 nanocomposites using ultrasonic nondestructive testing method.

scholarly journals VALIDATION OF A SALIVA-BASED TEST FOR THE MOLECULAR DIAGNOSIS OF SARS-CoV-2 INFECTION

2021 ◽  
Author(s):  
Michela Bulfoni ◽  
Emanuela Sozio ◽  
Barbara Marcon ◽  
Maria De Martino ◽  
Daniela Cesselli ◽  
...  
Keyword(s):  
Predictive Value ◽  
Saliva Sample ◽  
Scale Up ◽  
Rt Pcr ◽  
Clinical Sensitivity ◽  
Virus Identification ◽  
Stabilizing Solution ◽  
The Difference ◽  
Asymptomatic Subjects ◽  
Screening And Diagnosis

Background: Since the beginning of the pandemic, clinicians and researchers have been searching for alternative tests to improve screening and diagnosis of SARS-CoV-2 infection. Currently, the gold standard for virus identification is the nasopharyngeal (NP) swab. Saliva samples, however, offer clear practical and logistical advantages but due to lack of collection, transport, and storage solutions, high-throughput saliva-based laboratory tests are difficult to scale up as a screening or diagnostic tool. With this study, we aimed to validate an intra-laboratory molecular detection method for SARS-CoV-2 on saliva samples collected in a new storage and inactivating solution, comparing the results to NP swabs to determine the difference in sensitivity between the two tests. Methods: In this study, 156 patients (cases) and 1005 asymptomatic subjects (controls) were enrolled and tested simultaneously for the detection of the SARS-CoV-2 viral genome by RT-PCR on both NP swab and saliva samples. Saliva samples were collected in a preservative and inhibiting saline solution (Biofarma Srl). Internal method validation was performed to standardize the entire workflow for saliva samples. Results: The identification of SARS-CoV-2 conducted on saliva samples showed a clinical sensitivity of 95.1% and specificity of 97.8% compared to NP swabs. The positive predictive value (PPV) was 81% while the negative predictive value (NPV) was 99.5%. Test concordance was 97.6% (Cohen's Kappa=0.86; 95% CI 0.81-0.91). The LoD of the test was 5 viral copies for both samples. Conclusions: RT-PCR assays conducted on a stored saliva sample achieved similar performance to those on NP swabs and this may provide a very effective tool for population screening and diagnosis. Collection of saliva in a stabilizing solution makes the test more convenient and widely available; furthermore, the denaturation properties of the solution reduces the infective risks belonging to sample manipulation.

scholarly journals Pengaruh Kepemimpinan Dan Kompetensi Terhadap Produktivitas Pegawai PPSU (Studi Kasus Di Penanganan Prasarana Dan Sarana Umum Pada Kelurahan Lebak Bulus Jakarta Selatan)

2019 ◽  
Vol 2 (2) ◽  
pp. 189
Author(s):  
Rahmi Hermawati ◽  
Rizky Radhika Hidayat
Keyword(s):  
Regression Analysis ◽  
Sampling Method ◽  
Hypothesis Test ◽  
Test Validity ◽  
Collection Method ◽  
Urban Village ◽  
Sample Data ◽  
Leadership Competence ◽  
Multiple Analysis ◽  
Classical Regression

ABSTRACT As for the purpose of this study is to find and obtains evidence empirical and conclusions about influence style leadership and competence of the productivity employees infrastructure and facilities common (PPSU) in urban lebak bulus of south jakarta.            Research methodology used is associative with the quantitative approach. A method of the withdrawal of in the sample have used of these tests are the sampling method of saturated pt pgn promised to supply dating techniques in the entire household sample if all members of a percent of the population used as included in the research sample. Data collection method that we use is to a method of kuosioner and observation to the employees PPSU to officials in urban village lebak bulus of south jakarta . The method of analysis processing the data used was test validity, reabilitas test, test the assumption classical, regression analysis multiple, analysis koefisiensi determination ( r2 ), the hypothesis ( test f and t).            Based on the results of the analysis that has been done obtained that is the or relation a positive and welfare between style leadership and competence to productivity employees together or simultaneous is as much as r = 0,377, this showed that productivity employees at PPSU urban village lebak bulus south jakarta effected by force leadership and competence to a category strong contributing 37.7%, while or 62.3% influenced by other factors out variables tested in pengelitian this.The results of the regression equation is linear multiple namely y = 27,117 + 0,008X1 + 0,817X2 means if the force of leadership (X1) up 1 a unit of then will increase 0,008 and competence (X2) up 1 a unit of then will increase 0,817.Influence between leadership and competence to productivity employees is significant.This can be seen of the value of f count 25,972 > f table 3,10  Key Words : Leadership, Competence , Productivity

scholarly journals Peramalan Data Inflow dan Outflow Uang Kartal Bank Indonesia Provinsi DKI Jakarta Menggunakan Model ARIMAX dan SARIMAX

2021 ◽  
Vol 4 (2) ◽  
pp. 87
Author(s):  
Atika Amalia ◽  
Etik Zukhronah ◽  
Sri Subanti
Keyword(s):  
Time Series ◽  
Crucial Role ◽  
Forecasting Model ◽  
Gregorian Calendar ◽  
Monthly Data ◽  
Sample Data ◽  
The Difference

<p><strong>A</strong><strong>bstract</strong><strong>.</strong> DKI Jakarta Province plays a crucial role as the center of government and economy in Indonesia. The description of currency inflows and outflows is highly required before Bank Indonesia formulates the appropriate policies to control the circulation of money. The monthly data of currency inflow and outflow of Bank Indonesia of DKI Jakarta show a significant increase in each year particularly before, during, and after Eid al-Fitr. The determination of Eid al-Fitr does not follow the Gregorian calendar but based on the Islamic calendar. The difference in the use of the Gregorian and Islamic calendars in a time series causes a calendar variation. Thus, the determination of Eid al-Fitr in the Gregorian calendar changes as it goes forward eleven days each year or one month every three years. This study aims to obtain the best model and forecast currency inflows and outflows of Bank Indonesia DKI Jakarta using the ARIMAX and SARIMAX models. The study used in-sample data from January 2009 to December 2018 and out-sample data from January to October 2019. The best model was selected based on the smallest out-sample MAPE value. The result showed that the best forecasting model of inflow was ARIMAX (1,0,1). Meanwhile, the best forecasting model for outflow was SARIMAX (2,0,1)(0,0,1)<sup>12</sup>.</p><p><strong>Keywords: </strong>ARIMAX, calendar variation, forecasting, SARIMAX</p>

scholarly journals Perbandingan Metode Pembelajaran Number Heads Together (NHT) Dan Metode Pembelajaran Student Team Achievement Division (STAD) Terhadap Hasil Belajar Siswa Kelas VII Di MTs Madinatussalam

2021 ◽  
Vol 9 (3) ◽  
pp. 195-201
Author(s):  
Adha Siagian ◽  
◽  
Kartika Manalu ◽  
Khairuddin Khairuddin
Keyword(s):  
Learning Outcomes ◽  
Test Score ◽  
Hypothesis Test ◽  
Class Ii ◽  
Class I ◽  
Choice Test ◽  
Learning Method ◽  
High Category ◽  
Post Test ◽  
The Difference

This study aims to determine the differences between NHT) and STAD learning outcomes of clas VII Biology MTs Madinatussalam. The sample of this research is class VII-1 with 32 student as NHT class and 32 student in VII-2 as STAD class. The instrument used in this study was a multiple choice test concisting of 20 questions. The results of analysis showed that the average post-test score for the experimental class I NHT was 82 very high category. Meanwhile, the experimental class II STAD the average post-test score was 67,2 high category. The hypothesis test of the difference in learning outcomes of students in expremental class I NHT and expremental class II STAD, obtained tcount = 6,036>ttable = 1.999, then Ho is rejected and Ha is acepted. This shows that there is a difference in the biology lerning out comes of student who are thought using NHT learning method and those taught using the STAD learning method.

Sign in / Sign up

Export Citation Format

Share Document