Search

CoW using TSV technology has been suggested for process flexibility with high bandwidth in wafer level process (WLP). To connect I/O terminals, Thermo-compression process is generally used with using of non-conductive film (NCF). Although TC-NCF can suppress inner void and have a high production efficiency, solder sweep phenomenon, one of device fail reasons between signal bumps, was reported. This paper focus on design for three dimensional packaging technology and discuss optimal bonding process, NCF material and bump array design. TC-bonding simulation is also adapted for increasing experimental accuracy. The simulation is successfully estimated a sweep behavior and well matched with actual experimental results. It is expected that I/O counts per unit area is be expected to increase about 4 times and flexibility of I/O layout was also ensured.

Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is an unexplained chronic, debilitating illness characterized by fatigue, sleep disturbances, cognitive dysfunction, orthostatic intolerance and gastrointestinal problems. Using ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS), we analyzed the plasma proteomes of 39 ME/CFS patients and 41 healthy controls. Logistic regression models, with both linear and quadratic terms of the protein levels as independent variables, revealed a significant association between ME/CFS and the immunoglobulin heavy variable (IGHV) region 3-23/30. Stratifying the ME/CFS group based on self-reported irritable bowel syndrome (sr-IBS) status revealed a significant quadratic effect of immunoglobulin lambda constant region 7 on its association with ME/CFS with sr-IBS whilst IGHV3-23/30 and immunoglobulin kappa variable region 3-11 were significantly associated with ME/CFS without sr-IBS. In addition, we were able to predict ME/CFS status with a high degree of accuracy (AUC = 0.774-0.838) using a panel of proteins selected by 3 different machine learning algorithms: Lasso, Random Forests, and XGBoost. These algorithms also identified proteomic profiles that predicted the status of ME/CFS patients with sr-IBS (AUC = 0.806-0.846) and ME/CFS without sr-IBS (AUC = 0.754-0.780). Our findings are consistent with a significant association of ME/CFS with immune dysregulation and highlight the potential use of the plasma proteome as a source of biomarkers for disease.

This study simulated the performance of Cu-cored solder joints in microelectronic components subjected to the extreme thermal cycling conditions often encountered in the automobile industry by comparing the thermal cycling behavior of Cu-cored solder joints containing two different coating layers of Sn–3.0Ag and Sn–1.0In with that of a baseline Sn–3.0Ag–0.5Cu solder joint under a severe temperature cycling range of −55 to +150 °C. Both Cu-cored solder joints can be considered a potential solution to interconnects in microelectronic semiconductor packages used under harsh thermal conditions on account of their greater resistance to thermal stress caused by the severe temperature cycling than the baseline Sn–3.0Ag–0.5Cu solder joint.

Keystroke dynamics-based authentication (KDA) is to verify a user's identity using not only the password but also keystroke dynamics. With a small number of patterns available, data quality is of great importance in KDA applications. Recently, the authors have proposed to employ artificial rhythms and tempo cues to improve data quality: consistency and uniqueness of typing patterns. This paper examines whether improvement in uniqueness and consistency translates into improvement in authentication performance in real-world applications. In particular, we build various novelty detectors using typing patterns based on various strategies in which artificial rhythms and/or tempo cues are implemented. We show that artificial rhythms and tempo cues improve authentication accuracies and that they can be applicable in practical authentication systems.

Objective: Computerized continuous performance test (CPT) is a widely used diagnostic tool for attention-deficit hyperactivity disorder (ADHD). It measures the number of correctly detected stimuli as well as response times. Typically, when calculating a cut-off score for discriminating between normal and abnormal, only the normal children's data are collected. Then the average and standard deviation of each measure or variable is computed. If any of variables is larger than 2 sigma above the average, that child is diagnosed as abnormal. We will call this approach as ''T-score 70'' classifier. However, its performance has a lot to be desired due to a high false negative error. In order to improve the classification accuracy we propose to use novelty detection approaches for supporting ADHD diagnosis. Methods and material: Novelty detection is a model building framework where a classifier is constructed using only one class of training data and a new input pattern is classified according to its similarity to the training data. A total of eight novelty detectors are introduced and applied to our ADHD datasets collected from two modes of tests, visual and auditory. They are evaluated and compared with the T-score model on validation datasets in terms of false positive and negative error rates, and area under receiver operating characteristics curve (AuROC). Results: Experimental results show that the cut-off score of 70 is suboptimal which leads to a low false positive error but a very high false negative error. A few novelty detectors such as Parzen density estimators yield much more balanced classification performances. Moreover, most novelty detectors outperform the T-score method for most age groups statistically with a significance level of 1% in terms of AuROC. In particular, we recommend the Parzen and Gaussian density estimators, kernel principal component analysis, one-class support vector machine, and K-means clustering novelty detector which can improve upon the T-score method on average by at least 30% for the visual test and 40% for the auditory test. In addition, their performances are relatively stable over various parameter values as long as they are within reasonable ranges. Conclusion: The proposed novelty detection approaches can replace the T-score method which has been considered the ''gold standard'' for supporting ADHD diagnosis. Furthermore, they can be applied to other psychological tests where only normal data are available.

Response modeling has become a key factor to direct marketing. In general, there are two stages in response modeling. The first stage is to identify respondents from a customer database while the second stage is to estimate purchase amounts of the respondents. This paper focuses on the second stage where a regression, not a classification, problem is solved. Recently, several non-linear models based on machine learning such as support vector machines (SVM) have been applied to response modeling. However, there is a major difficulty. A typical training dataset for response modeling is so large that modeling takes very long, or, even worse, modeling may be impossible. Therefore, sampling methods have been usually employed in practice. However a sampled dataset usually leads to lower accuracy. In this paper, we employed an @e-tube based sampling for support vector regression (SVR) which leads to better accuracy than the random sampling method.

This paper proposes to use novelty detection approaches to alleviate the class imbalance in response modeling. Two novelty detectors, one-class support vector machine (1-SVM) and learning vector quantization for novelty detection (LVQ-ND), are compared with binary classifiers for a catalogue mailing task with DMEF4 dataset. The novelty detectors are more accurate and more profitable when the response rate is low. When the response rate is relatively high, however, a support vector machine model with modified misclassification costs performs the best. In addition, the novelty detectors turn in higher profits with a low mailing cost, while the SVM model is the most profitable with a high mailing cost.

In keystroke dynamics-based authentication, novelty detection methods are used since only the valid user's patterns are available when a classifier is first constructed. After a while, however, impostors' keystroke patterns become available from failed login attempts. We propose to employ the retraining framework where a novelty detector is retrained with the impostor patterns to enhance authentication accuracy. In this paper, learning vector quantization for novelty detection and support vector data description are retrained with the impostor patterns. Experimental results show that retraining improves the authentication performance and that learning vector quantization for novelty detection outperforms other widely used novelty detectors.

We propose to use learning vector quantization (LVQ) in novelty detection where a few outliers exist in training data. The codebook update of original LVQ is modified and the scheme to determine a threshold for each codebook is proposed. Experimental results on artificial and real-world problems are quite promising.

Crystal growth and haze formation on photomasks become serious problems in UV lithography. As the wavelength becomes shorter, photons carry more energy, so the chances of having a photochemical reaction become much higher. Pellicle, adhesive, residue from cleaning or resist strip process, and any contaminant in air can react with UV to form unwanted crystals and a haze layer on reticles. These will reduce the light transmission during exposure process. Thus, frequent mask inspection and periodic mask cleaning are needed to overcome these problems. However, these will in turn increase manufacturing cost and reduce mask life. Thus, a proper mask inspection tool is required to provide early warning of haze formation. In this work, we devised a new ellipsometric technique to investigate the inner surface of mask without removing pellicle. Ellipsometry is known to have mono-layer sensitivity and it can be used to measure any film or partial film formed on non-patterned spot in early stage of growth. However, when a pellicle covers the surface of mask, the ellipsometric data reflected from surface are extremely distorted due to the non-normal transmission through the pellicle. Thus, data analysis becomes extremely difficult without knowing the optical properties of pellicles. In order to solve this problem we developed compensation technique in which two blank pellicles are situated in the optical path in a way to compensate the polarization changes caused by the pellicle on mask. With this method, the conventional ellipsometry spectra of {Δ, Ψ} are deduced.

Keystroke dynamics-based authentication (KDA) is to verify a user's identification using not only the password but also keystroke patterns. The authors have shown in previous research that uniqueness and consistency of keystroke patterns are important factors to authentication accuracy and that they can be improved by employing artificial rhythms and tempo cues. In this paper, we implement the pause strategy and/or auditory cues for KDA and assess their effectiveness using various novelty detectors. Experimental results show that improved uniqueness and consistency lead to enhanced authentication performance, in particular for those users with poor typing ability.

In keystroke dynamics based user authentication, an access system utilizes not only a valid user's password, but also his/her typing patterns. Although high performances in terms of FAR(False Acceptance Rate) and FRR(False Rejection Rate) have been reported, most researches used a large number of valid users' typing patterns in order to implement complex algorithms in building a classifier[1]. However, collecting sufficient typing patterns to construct a complex classifier is practically impossible. When only a handful of typing patterns are available, the only way to compensate the lack of quantity is to improve quality. To improve the quality of typing patterns, using artificial rhythms and cues were proposed[2]. In this paper, we aim at verifying the effectiveness of artificial rhythms and cues by testing hypotheses.

Novelty detection involves identifying novel patterns. They are not usually available during training. Even if they are, the data quantity imbalance leads to a low classification accuracy when a supervised learning scheme is employed. Thus, an unsupervised learning scheme is often employed ignoring those few novel patterns. In this paper, we propose two ways to make use of the few available novel patterns. First, a scheme to determine local thresholds for the Self Organizing Map boundary is proposed. Second, a modification of the Learning Vector Quantization learning rule is proposed so that allows one to keep codebook vectors as far from novel patterns as possible. Experimental results are quite promising.

Observational learning algorithm is an ensemble algorithm where each network is initially trained with a bootstrapped data set and virtual data are generated from the ensemble for training. Here we propose a modular OLA approach where the original training set is partitioned into clusters and then each network is instead trained with one of the clusters. Networks are combined with different weighting factors now that are inversely proportional to the distance from the input vector to the cluster centers. Comparison with bagging and boosting shows that the proposed approach reduces generalization error with a smaller number of networks employed.