| 1. |
|
|
| 2. |
- Alabdallah, Abdallah, 1979-, et al.
(författare)
-
Discovering Premature Replacements in Predictive Maintenance Time-to-Event Data
- 2023
-
Ingår i: Proceedings of the Asia Pacific Conference of the PHM Society 2023. - New York : The Prognostics and Health Management Society.
-
Konferensbidrag (refereegranskat)abstract
- Time-To-Event (TTE) modeling using survival analysis in industrial settings faces the challenge of premature replacements of machine components, which leads to bias and errors in survival prediction. Typically, TTE survival data contains information about components and if they had failed or not up to a certain time. For failed components, the time is noted, and a failure is referred to as an event. A component that has not failed is denoted as censored. In industrial settings, in contrast to medical settings, there can be considerable uncertainty in an event; a component can be replaced before it fails to prevent operation stops or because maintenance staff believe that the component is faulty. This shows up as “no fault found” in warranty studies, where a significant proportion of replaced components may appear fault-free when tested or inspected after replacement.In this work, we propose an expectation-maximization-like method for discovering such premature replacements in survival data. The method is a two-phase iterative algorithm employing a genetic algorithm in the maximization phase to learn better event assignments on a validation set. The learned labels through iterations are accumulated and averaged to be used to initialize the following expectation phase. The assumption is that the more often the event is selected, the more likely it is to be an actual failure and not a “no fault found”.Experiments on synthesized and simulated data show that the proposed method can correctly detect a significant percentage of premature replacement cases.
|
|
| 3. |
- Alabdallah, Abdallah, 1979-
(författare)
-
Machine Learning Survival Models : Performance and Explainability
- 2023
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Survival analysis is an essential statistics and machine learning field in various critical applications like medical research and predictive maintenance. In these domains understanding models' predictions is paramount. While machine learning techniques are increasingly applied to enhance the predictive performance of survival models, they simultaneously sacrifice transparency and explainability. Survival models, in contrast to regular machine learning models, predict functions rather than point estimates like regression and classification models. This creates a challenge regarding explaining such models using the known off-the-shelf machine learning explanation techniques, like Shapley Values, Counterfactual examples, and others. Censoring is also a major issue in survival analysis where the target time variable is not fully observed for all subjects. Moreover, in predictive maintenance settings, recorded events do not always map to actual failures, where some components could be replaced because it is considered faulty or about to fail in the future based on an expert's opinion. Censoring and noisy labels create problems in terms of modeling and evaluation that require to be addressed during the development and evaluation of the survival models.Considering the challenges in survival modeling and the differences from regular machine learning models, this thesis aims to bridge this gap by facilitating the use of machine learning explanation methods to produce plausible and actionable explanations for survival models. It also aims to enhance survival modeling and evaluation revealing a better insight into the differences among the compared survival models.In this thesis, we propose two methods for explaining survival models which rely on discovering survival patterns in the model's predictions that group the studied subjects into significantly different survival groups. Each pattern reflects a specific survival behavior common to all the subjects in their respective group. We utilize these patterns to explain the predictions of the studied model in two ways. In the first, we employ a classification proxy model that can capture the relationship between the descriptive features of subjects and the learned survival patterns. Explaining such a proxy model using Shapley Values provides insights into the feature attribution of belonging to a specific survival pattern. In the second method, we addressed the "what if?" question by generating plausible and actionable counterfactual examples that would change the predicted pattern of the studied subject. Such counterfactual examples provide insights into actionable changes required to enhance the survivability of subjects.We also propose a variational-inference-based generative model for estimating the time-to-event distribution. The model relies on a regression-based loss function with the ability to handle censored cases. It also relies on sampling for estimating the conditional probability of event times. Moreover, we propose a decomposition of the C-index into a weighted harmonic average of two quantities, the concordance among the observed events and the concordance between observed and censored cases. These two quantities, weighted by a factor representing the balance between the two, can reveal differences between survival models previously unseen using only the total Concordance index. This can give insight into the performances of different models and their relation to the characteristics of the studied data.Finally, as part of enhancing survival modeling, we propose an algorithm that can correct erroneous event labels in predictive maintenance time-to-event data. we adopt an expectation-maximization-like approach utilizing a genetic algorithm to find better labels that would maximize the survival model's performance. Over iteration, the algorithm builds confidence about events' assignments which improves the search in the following iterations until convergence.We performed experiments on real and synthetic data showing that our proposed methods enhance the performance in survival modeling and can reveal the underlying factors contributing to the explainability of survival models' behavior and performance.
|
|
| 4. |
- Alabdallah, Abdallah, 1979-, et al.
(författare)
-
SurvSHAP : A Proxy-Based Algorithm for Explaining Survival Models with SHAP
- 2022
-
Ingår i: 2022 IEEE 9th International Conference on Data Science and Advanced Analytics (DSAA). - Piscataway, NJ : IEEE. - 9781665473309 - 9781665473316
-
Konferensbidrag (refereegranskat)abstract
- Survival Analysis models usually output functions (survival or hazard functions) rather than point predictions like regression and classification models. This makes the explanations of such models a challenging task, especially using the Shapley values. We propose SurvSHAP, a new model-agnostic algorithm to explain survival models that predict survival curves. The algorithm is based on discovering patterns in the predicted survival curves, the output of the survival model, that would identify significantly different survival behaviors, and utilizing a proxy model and SHAP method to explain these distinct survival behaviors. Experiments on synthetic and real datasets demonstrate that the SurvSHAP is able to capture the underlying factors of the survival patterns. Moreover, SurvSHAP results on the Cox Proportional Hazard model are compared with the weights of the model to show that we provide faithful overall explanations, with more fine-grained explanations of the sub-populations. We also illustrate the wrong model and explanations learned by a Cox model when applied to heterogeneous sub-populations. We show that a non-linear machine learning survival model with SurvSHAP can better model the data and provide better explanations than linear models.
|
|
| 5. |
- Alabdallah, Abdallah, 1979-, et al.
(författare)
-
The Concordance Index decomposition : A measure for a deeper understanding of survival prediction models
- 2024
-
Ingår i: Artificial Intelligence in Medicine. - Amsterdam : Elsevier B.V.. - 0933-3657 .- 1873-2860. ; 148
-
Tidskriftsartikel (refereegranskat)abstract
- The Concordance Index (C-index) is a commonly used metric in Survival Analysis for evaluating the performance of a prediction model. In this paper, we propose a decomposition of the C-index into a weighted harmonic mean of two quantities: one for ranking observed events versus other observed events, and the other for ranking observed events versus censored cases. This decomposition enables a finer-grained analysis of the relative strengths and weaknesses between different survival prediction methods. The usefulness of this decomposition is demonstrated through benchmark comparisons against classical models and state-of-the-art methods, together with the new variational generative neural-network-based method (SurVED) proposed in this paper. The performance of the models is assessed using four publicly available datasets with varying levels of censoring. Using the C-index decomposition and synthetic censoring, the analysis shows that deep learning models utilize the observed events more effectively than other models. This allows them to keep a stable C-index in different censoring levels. In contrast to such deep learning methods, classical machine learning models deteriorate when the censoring level decreases due to their inability to improve on ranking the events versus other events.
|
|
| 6. |
- Alabdallah, Abdallah, 1979-, et al.
(författare)
-
Understanding Survival Models through Counterfactual Explanations
-
Annan publikation (övrigt vetenskapligt/konstnärligt)abstract
- The development of black-box survival models has created a need for methods that explain their outputs, just as in the case of traditional machine learning methods. Survival models usually predict functions rather than point estimates. This special nature of their output makes it more difficult to explain their operation. We propose a method to generate plausible counterfactual explanations for survival models. The method supports two options that handle the special nature of survival models' output. One option relies on the Survival Scores, which are based on the area under the survival function, which is more suitable for proportional hazard models. The other one relies on Survival Patterns in the predictions of the survival model, which represent groups that are significantly different from the survival perspective. This guarantees an intuitive well-defined change from one risk group (Survival Pattern) to another and can handle more realistic cases where the proportional hazard assumption does not hold. The method uses a Particle Swarm Optimization algorithm to optimize a loss function to achieve four objectives: the desired change in the target, proximity to the explained example, likelihood, and the actionability of the counterfactual example. Two predictive maintenance datasets and one medical dataset are used to illustrate the results in different settings. The results show that our method produces plausible counterfactuals, which increase the understanding of black-box survival models.
|
|
| 7. |
- Ali Hamad, Rebeen, 1989-
(författare)
-
Towards Reliable, Stable and Fast Learning for Smart Home Activity Recognition
- 2022
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The current population age grows increasingly in industrialized societies and calls for more intelligent tools to monitor human activities. The aims of these intelligent tools are often to support senior people in their homes, to keep track of their daily activities, and to early detect potential health problems to facilitate a long and independent life. The recent advancements of smart environments using miniaturized sensors and wireless communications have facilitated unobtrusively human activity recognition. Human activity recognition has been an active field of research due to its broad applications in different areas such as healthcare and smart home monitoring. This thesis project develops work on machine learning systems to improve the understanding of human activity patterns in smart home environments. One of the contributions of this research is to process and share information across multiple smart homes to reduce the learning time, reduce the need and effort to recollect the training data, as well as increase the accuracy for applications such as activity recognition. To achieve that, several contributions are presented to pave the way to transfer knowledge among smart homes that includes the following studies. Firstly, a method to align manifolds is proposed to facilitate transfer learning. Secondly, we propose a method to further improve the performance of activity recognition over the existing methods. Moreover, we explore imbalanced class problems in human activity recognition and propose a method to handle imbalanced human activities. The summary of these studies are provided below. In our work, it is hypothesized that aligning learned low-dimensional manifolds from disparate datasets could be used to transfer knowledge between different but related datasets. The t-distributed Stochastic Neighbor Embedding(t-SNE) is used to project the high-dimensional input dataset into low-dimensional manifolds. However, since t-SNE is a stochastic algorithm and there is a large variance of t-SNE maps, a thorough analysis of the stability is required before applying Transfer learning. In response to this, an extension to Local Procrustes Analysis called Normalized Local Procrustes Analysis (NLPA) is proposed to non-linearly align manifolds by using locally linear mappings to test the stability of t-SNE low-dimensional manifolds. Experiments show that the disparity from using NLPA to align low-dimensional manifolds decreases by order of magnitude compared to the disparity obtained by Procrustes Analysis (PA). NLPA outperforms PA and provides much better alignments for the low-dimensional manifolds. This indicates that t-SNE low-dimensional manifolds are locally stable, which is the part of the contribution in this thesis.Human activity recognition in smart homes shows satisfying recognition results using existing methods. Often these methods process sensor readings that precede the evaluation time (where the decision is made) to evaluate and deliver real-time human activity recognition. However, there are several critical situations, such as diagnosing people with dementia where "preceding sensor activations" are not always sufficient to accurately recognize the resident's daily activities in each evaluated time. To improve performance, we propose a method that delays the recognition process to include some sensor activations that occur after the point in time where the decision needs to be made. For this, the proposed method uses multiple incremental fuzzy temporal windows to extract features from both preceding and some oncoming sensor activations. The proposed method is evaluated with two temporal deep learning models: one-dimensional convolutional neural network (1D CNN) and long short-term memory (LSTM) on a binary sensor dataset of real daily living activities. The experimental evaluation shows that the proposed method achieves significantly better results than the previous state-of-the-art. Further, one of the main problems of activity recognition in a smart home setting is that the frequency and duration of human activities are intrinsically imbalanced. The huge difference in the number of observations for the categories means that many machine learning algorithms focus on the classification of the majority examples due to their increased prior probability while ignoring or misclassifying minority examples. This thesis explores well-known class imbalance approaches (synthetic minority over-sampling technique, cost-sensitive learning and ensemble learning) applied to activity recognition data with two temporal data pre-processing for the deep learning models LSTM and 1D CNN. This thesis proposes a data level perspective combined with a temporal window technique to handle imbalanced human activities from smart homes in order to make the learning algorithms more sensitive to the minority class. The experimental results indicate that handling imbalanced human activities from the data-level outperforms algorithm level and improved the classification performance.
|
|
| 8. |
- Altarabichi, Mohammed Ghaith, 1981-, et al.
(författare)
-
Improving Concordance Index in Regression-based Survival Analysis : Discovery of Loss Function for Neural Networks
- 2024
-
Annan publikation (övrigt vetenskapligt/konstnärligt)abstract
- In this work, we use an Evolutionary Algorithm (EA) to discover a novel Neural Network (NN) regression-based survival loss function with the aim of improving the C-index performance. Our contribution is threefold; firstly, we propose an evolutionary meta-learning algorithm SAGA$_{loss}$ for optimizing a neural-network regression-based loss function that maximizes the C-index; our algorithm consistently discovers specialized loss functions that outperform MSCE. Secondly, based on our analysis of the evolutionary search results, we highlight a non-intuitive insight that signifies the importance of the non-zero gradient for the censored cases part of the loss function, a property that is shown to be useful in improving concordance. Finally, based on this insight, we propose MSCE$_{Sp}$, a novel survival regression loss function that can be used off-the-shelf and generally performs better than the Mean Squared Error for censored cases. We performed extensive experiments on 19 benchmark datasets to validate our findings.
|
|
| 9. |
- Altarabichi, Mohammed Ghaith, 1981-, et al.
(författare)
-
Predicting state of health and end of life for batteries in hybrid energy buses
- 2020
-
Ingår i: Proceedings of the 30th European Safety and Reliability Conference and the 15th Probabilistic Safety Assessment and Management Conference. - Singapore : Research Publishing Services. - 9789811485930 ; , s. 1231-1231
-
Konferensbidrag (refereegranskat)abstract
- There is a major ongoing transition from utilizing fossil fuel to electricity in buses for enabling a more sustainable, environmentally friendly, and connected transportation ecosystem. Batteries are expensive, up to 30% of the total cost for the vehicle (A. Fotouhi 2016), and considered safety-critical components for electric vehicles (EV). As they deteriorate over time, monitoring the health status and performing the maintenance accordingly in a proactive manner is crucial to achieving not only a safe and sustainable transportation system but also a cost-effective operation and thus a greater market satisfaction. As a widely used indicator, the State of Health (SOH) is a measurement that reflects the current capability of the battery in comparison to an ideal condition. Accurate estimation of SOH is important to evaluate the validity of the batteries for the intended application and can be utilized as a proxy to estimate the remaining useful life (RUL) and predict the end-of-life (EOL) of batteries for maintenance planning. The SOH is computed via an on-board computing device, i.e. battery management unit (BMU), which is commonly developed based on controlled experiments and many of them are physical-model based approaches that only depend on the internal parameters of the battery (B. Pattipati 2008; M. H. Lipu 2018). However, the deterioration processes of batteries in hybrid and full-electric buses depend not only on the designing parameters but also on the operating environment and usage patterns of the vehicle. Therefore, utilizing multiple data sources to estimate the health status and EOL of the batteries is of potential internet. In this study, a data-driven prognostic method is developed to estimate SOH and predict EOL for batteries in heterogeneous fleets of hybrid buses, using various types of data sources, e.g. physical configuration of the vehicle, deployment information, on-board sensor readings, and diagnostic fault codes. A set of new features was generated from the existing sensor readings by inducing artificial resets on each battery replacement. A neural network-based regression model achieved accurate estimates of battery SOH status. Another network was used to indicate the EOL of batteries and the result was evaluated using battery replacement based on the current maintenance strategy. © ESREL2020-PSAM15 Organizers. Published by Research Publishing, Singapore.
|
|
| 10. |
- Altarabichi, Mohammed Ghaith, 1981-, et al.
(författare)
-
Stacking Ensembles of Heterogenous Classifiers for Fault Detection in Evolving Environments
- 2020
-
Ingår i: Proceedings of the 30th European Safety and Reliability Conference and the 15th Probabilistic Safety Assessment and Management Conference. - Singapore : Research Publishing Services. - 9789811485930 ; , s. 1068-1068
-
Konferensbidrag (refereegranskat)abstract
- Monitoring the condition, detecting faults, and modeling the degradation of industrial equipment are important challenges in Prognostics and Health Management (PHM) field. Our solution to the challenge demonstrated a multi-stage approach for detecting faults in a group of identical industrial equipment, composed of four identical interconnected components, that have been deployed to the evolving environment with changes in operational and environmental conditions. In the first stage, a stacked ensemble of heterogeneous classifiers was applied to predict the state of each component of the equipment individually. In the second stage, a low pass filter was applied to smoothen the predictions cast by stacked ensembles, utilizing temporal information of the prediction sequence. © ESREL2020-PSAM15 Organizers. Published by Research Publishing, Singapore.
|
|
