Bayesian network for integrated circuit testing probe card fault diagnosis and troubleshooting to empower Industry 3.5 smart production and an empirical study

Automatic Design of a Broadband Directional Coupler via Bayesian Optimization

“…This problem is also known as circuit synthesis or circuit inverse design in the literature. Fault diagnosis [52] Belief network Contact resistance Results of 3 logged trails matched with the expected [53] BNN Differential amplifier < 1% prediction error over 60% test cases [185] BNN + preprocessing Analog filter 96% correct classification of test data [48] Bayes' theorem Two-stage Op-Amp 86% of process/layout faults diagnosed [186] Belief network ISCAS'89 benchmark circuits Transient error diagnosed [187] Belief network Chipset assembly, test factory Good agreement with real data [188] Belief network Etch equipment 4 out of 5 faults identified correctly [189] Belief network A wafer dataset Over 90% accuracy in most cases [190] Gibbs sampling 20 lots of 500 wafers Smaller error compared to baselines [191] Variational Bayes Chaotic circuit, Op-Amp filter > 1% ↑ accuracy compared to baselines [192] Naive Bayes classifier Chaotic circuit, Op-Amp filter 0.2%, 2.5%, 3.2% ↑ accuracy compared to baselines [193] Belief network Probe card About 3× accuracy improvement and 64% cost reduction Reliability analysis [184] MAP A lithography tool Test length shortened [194] PTM logic circuits > 5% ↑ reliability improved [195] Bayes' theorem 1 logic circuits < 12% ↓ error compared to MC w/ less run-time [196] Belief network CMOS gates Better accuracy compared with one baseline [197] Belief network 5 logic circuits > 2% ↑ accuracy compared to baselines [198] MAP HV circuit breaker Reliability value obtained [199] Belief network 2 ISCAS'85 benchmark circuits Faster and more accurate compared to exact inference [200] Belief network ZPW-2000A track circuits Reliability value obtained [201] Belief network Microprocessors More accurate FIT compared to FI w/ 65% ↓ run-time [202] Belief network COTS devices Failure rate calculated Process control [55] Empirical Bayes Printed circuit board Deployed in AT&T factories [54] MAP VLSIC process Detection time reduced for a multi-stage process [203] Particle filter CMP process (MRR prediction) 13% ↑ improvement compared to baselines [204] Particle filter CMP process (end-po...…”

Application of machine learning based models in computer network data

“…In network management, troubleshooting is one of the important and difficult tasks [1][2]. Fault diagnosis is a process of finding the root cause of a fault based on the observed events and generally consists of three steps: fault detection, fault location, and testing [3][4]. Fault detection determines whether a fault exists in the network based on the signs of failure.…”

Research on Financing Environment Evaluation of Scientific Innovation Industry Based on the Bayesian Network Model under the Background of Green Economy

“…By applying the Bayesian network method, domestic and foreign scholars have made certain research achievements. Fu et al applied it to fault diagnosis and elimination of integrated circuit test probes [25], Wang et al applied it to scenario analysis under the cognitive uncertainty of scientific and technological accidents [26], and Cao applied it to personal credit evaluation. To predict the default risk of future borrowers [27], Wang et al applied it to predict the probability of technological progress to get the path of technological optimization [28], Li et al combined it with the explanatory structure model to build the risk identification and early warning model of social media network public opinion under emergencies [29], Bai et al combined it with the rough set theory.…”