Israel Beniaminy - Academia.edu (original) (raw)
Uploads
Papers by Israel Beniaminy
In discussing diagnostic expert systems, there is an ongoing debate as to whether model-based sys... more In discussing diagnostic expert systems, there is an ongoing debate as to whether model-based systems are superior to case-based systems, or vice versa. Our experience has shown that there is no real need for debate because the two are not mutually exclusive and, to the contrary, complement each other. Current expert system technology is capable of two reasoning mechanisms, in
La presente invention concerne un procede permettant de partager des connaissances entre une plur... more La presente invention concerne un procede permettant de partager des connaissances entre une pluralite de personnes. Cette invention comprend une base de connaissances mise a jour de facon dynamique qui peut etre accedee via un reseau. Une personne peut rencontrer un probleme dont la solution n'est pas connue et questionner cette base de donnee de facon a determiner si ce probleme et sa solution s'y trouvent. Si la solution de ce probleme ne se trouve pas dans cette base de donnee, la personne peut preparer une demande comprenant les details de ce probleme. Cette demande est diffusee via le reseau, et les experts presents sur le reseau peuvent repondre a cette demande sur le formulaire de demande, et s'ils le souhaitent, ces experts peuvent repondre par une communication personnelle. La reponse a cette question est acheminee via le reseau a la personne qui a interroge le systeme. Lorsque ce probleme a ete resolu, ou lorsque les suggestions n'ont pas donne de resultat...
Studies in Computational Intelligence
We discuss a class of large-scale real-world field service optimization problems which may be des... more We discuss a class of large-scale real-world field service optimization problems which may be described as generalizations of the Vehicle Routing Problem with Time Windows (VRPTW). We describe our experience in the real-world issues concerned with describing and solving instances of such problems, and adapting the solution to the needs of service organizations using a ”universal framework” for bringing together
INFORMS Journal on Applied Analytics
We present an optimization-based decision support system to generate optimal aircraft engine main... more We present an optimization-based decision support system to generate optimal aircraft engine maintenance schedules that reflect qualitative and quantitative trade-offs from customer, business, and shop perspectives. The approach is currently implemented at GE Aviation Services for global overhaul network induction planning for all commercial product lines.
: AITEST is a real life expert system designed to serve as a decision aid and productivity tool f... more : AITEST is a real life expert system designed to serve as a decision aid and productivity tool for test engineers and technicians. Oriented for the functional level, AITEST is designed to troubleshoot large scale UUT's (Unit Under Test) that contain analog, digital and mechanical modules in electronic, electro optic, hydraulic or mechanical systems and devices. This paper describes a typical application of AITEST in an intermediate - level maintenance facility. The UUT (Unit Under Test) discussed in this application note is a Radar Modulator (RM) embedded in the Radar System of a military aircraft. Following a brief description of the Radar Modulator structure and functions and its current test tools, we proceed to describing our experience in using AITEST to troubleshoot this device.
ArXiv, 2020
Motion artefacts created by patient motion during an MRI scan occur frequently in practice, often... more Motion artefacts created by patient motion during an MRI scan occur frequently in practice, often rendering the scans clinically unusable and requiring a re-scan. While many methods have been employed to ameliorate the effects of patient motion, these often fall short in practice. In this paper we propose a novel method for removing motion artefacts using a deep neural network with two input branches that discriminates between patient poses using the motion's timing. The first branch receives a subset of the kkk-space data collected during a single patient pose, and the second branch receives the remaining part of the collected kkk-space data. The proposed method can be applied to artefacts generated by multiple movements of the patient. Furthermore, it can be used to correct motion for the case where kkk-space has been under-sampled, to shorten the scan time, as is common when using methods such as parallel imaging or compressed sensing. Experimental results on both simulated a...
Increasingly complex test equipment, complex technologies, growing pressure on technicians’ and e... more Increasingly complex test equipment, complex technologies, growing pressure on technicians’ and engineers’ performance, and the need for integrated knowledge during the problemresolution process, are all demanding a higher level of expertise and a new approach to training. CBT (Computer Based Training) allows trainees to study current available knowledge using a software format. Training becomes less expensive and more flexible in timing, level of expertise and location. For CBT, the most pressing issue is how to prepare and expedite indepth knowledge to CBT in the most optimal way. In this paper we describe the ICBT (Intelligent Computer Based Training) concept. ICBT is based on the technology of expert system logic, a combination of an expert system used for fault isolation and diagnosis and an interactive documentation tool. The trainee learns the process of diagnosis and fault isolation for test and measurement equipment without the need for extensive, and expensive, hands-on ex...
Motion artefacts created by patient motion during an MRI scan occur frequently in practice, often... more Motion artefacts created by patient motion during an MRI scan occur frequently in practice, often rendering the scans clinically unusable and requiring a re-scan. While many methods have been employed to ameliorate the effects of patient motion, these often fall short in practice. In this paper we propose a novel method for detecting and timing patient motion during an MR scan and correcting for the motion artefacts using a deep neural network. The deep neural network contains two input branches that discriminate between patient poses using the motion’s timing. The first branch receives a subset of the k-space data collected during a single dominant patient pose, and the second branch receives the remaining part of the collected k-space data. The proposed method can be applied to artefacts generated by multiple movements of the patient. Furthermore, it can be used to correct motion for the case where k-space has been under-sampled to shorten the scan time, as is common when using me...
Proceedings Ieee Aerospace Conference, 2002
1997 Ieee Autotestcon Proceedings Autotestcon 97 Ieee Systems Readiness Technology Conference Systems Readiness Supporting Global Needs and Awareness in the 21st Century, Sep 22, 1997
Proceedings Ieee Aerospace Conference, 2002
Lecture Notes in Computer Science, 2007
We consider the Generalized Scheduling Within Intervals (GSWI) problem: given a set J of jobs and... more We consider the Generalized Scheduling Within Intervals (GSWI) problem: given a set J of jobs and a set I of intervals, where each job j ∈ J has in interval I ∈ I length (processing time) j,I and profit p j,I , find the highest-profit feasible schedule. The best approximation ratio known for GSWI is (1/2 − ε). We give a (1 − 1/e − ε)approximation scheme for GSWI with bounded profits, based on the work by Chuzhoy, Rabani, and Ostrovsky [5], for the {0, 1}-profit case. We also consider the Scheduling Within Intervals (SWI) problem, which is a particular case of GSWI where for every j ∈ J there is a unique interval I = I j ∈ I with p j,I > 0. We prove that SWI is (weakly) NP-hard even if the stretch factor (the maximum ratio of job's interval size to its processing time) is arbitrarily small, and give a polynomial-time algorithm for bounded profits and stretch factor < 2.
In discussing diagnostic expert systems, there is an ongoing debate as to whether model-based sys... more In discussing diagnostic expert systems, there is an ongoing debate as to whether model-based systems are superior to case-based systems, or vice versa. Our experience has shown that there is no real need for debate because the two are not mutually exclusive and, to the contrary, complement each other. Current expert system technology is capable of two reasoning mechanisms, in
La presente invention concerne un procede permettant de partager des connaissances entre une plur... more La presente invention concerne un procede permettant de partager des connaissances entre une pluralite de personnes. Cette invention comprend une base de connaissances mise a jour de facon dynamique qui peut etre accedee via un reseau. Une personne peut rencontrer un probleme dont la solution n'est pas connue et questionner cette base de donnee de facon a determiner si ce probleme et sa solution s'y trouvent. Si la solution de ce probleme ne se trouve pas dans cette base de donnee, la personne peut preparer une demande comprenant les details de ce probleme. Cette demande est diffusee via le reseau, et les experts presents sur le reseau peuvent repondre a cette demande sur le formulaire de demande, et s'ils le souhaitent, ces experts peuvent repondre par une communication personnelle. La reponse a cette question est acheminee via le reseau a la personne qui a interroge le systeme. Lorsque ce probleme a ete resolu, ou lorsque les suggestions n'ont pas donne de resultat...
Studies in Computational Intelligence
We discuss a class of large-scale real-world field service optimization problems which may be des... more We discuss a class of large-scale real-world field service optimization problems which may be described as generalizations of the Vehicle Routing Problem with Time Windows (VRPTW). We describe our experience in the real-world issues concerned with describing and solving instances of such problems, and adapting the solution to the needs of service organizations using a ”universal framework” for bringing together
INFORMS Journal on Applied Analytics
We present an optimization-based decision support system to generate optimal aircraft engine main... more We present an optimization-based decision support system to generate optimal aircraft engine maintenance schedules that reflect qualitative and quantitative trade-offs from customer, business, and shop perspectives. The approach is currently implemented at GE Aviation Services for global overhaul network induction planning for all commercial product lines.
: AITEST is a real life expert system designed to serve as a decision aid and productivity tool f... more : AITEST is a real life expert system designed to serve as a decision aid and productivity tool for test engineers and technicians. Oriented for the functional level, AITEST is designed to troubleshoot large scale UUT's (Unit Under Test) that contain analog, digital and mechanical modules in electronic, electro optic, hydraulic or mechanical systems and devices. This paper describes a typical application of AITEST in an intermediate - level maintenance facility. The UUT (Unit Under Test) discussed in this application note is a Radar Modulator (RM) embedded in the Radar System of a military aircraft. Following a brief description of the Radar Modulator structure and functions and its current test tools, we proceed to describing our experience in using AITEST to troubleshoot this device.
ArXiv, 2020
Motion artefacts created by patient motion during an MRI scan occur frequently in practice, often... more Motion artefacts created by patient motion during an MRI scan occur frequently in practice, often rendering the scans clinically unusable and requiring a re-scan. While many methods have been employed to ameliorate the effects of patient motion, these often fall short in practice. In this paper we propose a novel method for removing motion artefacts using a deep neural network with two input branches that discriminates between patient poses using the motion's timing. The first branch receives a subset of the kkk-space data collected during a single patient pose, and the second branch receives the remaining part of the collected kkk-space data. The proposed method can be applied to artefacts generated by multiple movements of the patient. Furthermore, it can be used to correct motion for the case where kkk-space has been under-sampled, to shorten the scan time, as is common when using methods such as parallel imaging or compressed sensing. Experimental results on both simulated a...
Increasingly complex test equipment, complex technologies, growing pressure on technicians’ and e... more Increasingly complex test equipment, complex technologies, growing pressure on technicians’ and engineers’ performance, and the need for integrated knowledge during the problemresolution process, are all demanding a higher level of expertise and a new approach to training. CBT (Computer Based Training) allows trainees to study current available knowledge using a software format. Training becomes less expensive and more flexible in timing, level of expertise and location. For CBT, the most pressing issue is how to prepare and expedite indepth knowledge to CBT in the most optimal way. In this paper we describe the ICBT (Intelligent Computer Based Training) concept. ICBT is based on the technology of expert system logic, a combination of an expert system used for fault isolation and diagnosis and an interactive documentation tool. The trainee learns the process of diagnosis and fault isolation for test and measurement equipment without the need for extensive, and expensive, hands-on ex...
Motion artefacts created by patient motion during an MRI scan occur frequently in practice, often... more Motion artefacts created by patient motion during an MRI scan occur frequently in practice, often rendering the scans clinically unusable and requiring a re-scan. While many methods have been employed to ameliorate the effects of patient motion, these often fall short in practice. In this paper we propose a novel method for detecting and timing patient motion during an MR scan and correcting for the motion artefacts using a deep neural network. The deep neural network contains two input branches that discriminate between patient poses using the motion’s timing. The first branch receives a subset of the k-space data collected during a single dominant patient pose, and the second branch receives the remaining part of the collected k-space data. The proposed method can be applied to artefacts generated by multiple movements of the patient. Furthermore, it can be used to correct motion for the case where k-space has been under-sampled to shorten the scan time, as is common when using me...
Proceedings Ieee Aerospace Conference, 2002
1997 Ieee Autotestcon Proceedings Autotestcon 97 Ieee Systems Readiness Technology Conference Systems Readiness Supporting Global Needs and Awareness in the 21st Century, Sep 22, 1997
Proceedings Ieee Aerospace Conference, 2002
Lecture Notes in Computer Science, 2007
We consider the Generalized Scheduling Within Intervals (GSWI) problem: given a set J of jobs and... more We consider the Generalized Scheduling Within Intervals (GSWI) problem: given a set J of jobs and a set I of intervals, where each job j ∈ J has in interval I ∈ I length (processing time) j,I and profit p j,I , find the highest-profit feasible schedule. The best approximation ratio known for GSWI is (1/2 − ε). We give a (1 − 1/e − ε)approximation scheme for GSWI with bounded profits, based on the work by Chuzhoy, Rabani, and Ostrovsky [5], for the {0, 1}-profit case. We also consider the Scheduling Within Intervals (SWI) problem, which is a particular case of GSWI where for every j ∈ J there is a unique interval I = I j ∈ I with p j,I > 0. We prove that SWI is (weakly) NP-hard even if the stretch factor (the maximum ratio of job's interval size to its processing time) is arbitrarily small, and give a polynomial-time algorithm for bounded profits and stretch factor < 2.