Hyperparameter Configurations Research Articles

Minimizing Energy Cost in PV Battery Storage Systems Using Reinforcement Learning

This article addresses the development and tuning of an energy management for a photovoltaic (PV) battery storage system for the cost-optimized use of PV energy using of reinforcement learning (RL). An energy management concept based on the Proximal Policy Optimization algorithm in combination with recurrent Long Short-Term Memory neural networks is developed for data-based policy learning. As a reference system for the simulation-based investigations, a PV battery storage system is modelled, parametrized and implemented with an interface for the RL algorithm. To demonstrate the generalization capability of the learned energy management, 98 training and 12 evaluation episodes, each with a length of one year, are generated from an empirical dataset of global radiation and load power time series. To improve the convergence speed and stability of the RL algorithm as well as the learned policy with regards to techno-economic metrics, an extensive hyperparameter study is conducted by training 216 control policies with different hyperparameter configurations. Simulation-based benchmark tests of the learned energy management against conventional rule-based and model-predictive energy managements show that the RL-based concept can achieve slightly better results in terms of energy costs and the amount of energy fed into the grid than the commonly used model-predictive method.

On Hyperparameter Optimization of Machine Learning Methods Using a Bayesian Optimization Algorithm to Predict Work Travel Mode Choice

Prediction of work Travel mode choice is one of the most important parts of travel demand forecasting. Planners can achieve sustainability goals by accurately forecasting how people will get to and from work. In the prediction of travel mode selection, machine learning methods are commonly employed. To fit a machine-learning model to various challenges, the hyperparameters must be tweaked. Choosing the optimal hyperparameter configuration for machine learning models has an immediate effect on the performance of the model. In this paper, optimizing the hyperparameters of common machine learning models, including support vector machines, k-nearest neighbor, single decision trees, ensemble decision trees, and Naive Bayes, is studied using the Bayesian Optimization algorithm. These models were developed and optimized using two datasets from the 2017 National Household Travel Survey. Using several criteria, including average accuracy (%), average area under the receiver operating characteristics, and a simple ranking system, the performance of the optimized models was investigated. The findings of this study show that the BO is an effective model for improving the performance of the k-nearest neighbor model more than other models. This research lays the groundwork for using optimized machine learning methods to mitigate the negative consequences of automobile use.

Spatial rain probabilistic prediction performance using costsensitive learning algorithm

The use of machine learning in weather prediction is growing rapidly as an alternative to conventional numerical weather prediction. However, predictions using machine learning such as Long Short Term Memory (LSTM) based on neural networks have weaknesses in predicting extreme events with a high ratio of unbalanced data. This research examines the performance of using focal loss in LSTM to obtain a machine-learning model that is cost-sensitive. The model used the Global Forecasting System Data and the Global Satellite Measurement of Precipitation for the years 2017-2020. Testing the hyperparameter configuration was carried out using the hyperband method on the number of nodes and the number of iterations with 3 scenarios (2, 3, and 4 classes). The results showed an increased performance against noncost sensitive LSTM with an average increase of 25% accuracy and 11% F1-score on 2 classes scenario, 15% accuracy increase and 21% F1-score for scenario 3 classes, as well as an increase in accuracy of 15% and F1-score 26% for scenario 4 class. It also provides the idea of how cost-sensitive properties can help machine learning models detect classes with extreme ratios, based on an increase in average performance as the number of classification scenarios increases.

A Population-Based Hybrid Approach for Hyperparameter Optimization of Neural Networks

In recent years, large amounts of data have been generated, and computer power has kept growing. This scenario has led to a resurgence in the interest in artificial neural networks. One of the main challenges in training effective neural network models is finding the right combination of hyperparameters to be used. Indeed, the choice of an adequate approach to search the hyperparameter space directly influences the accuracy of the resulting neural network model. Common approaches for hyperparameter optimization are Grid Search, Random Search, and Bayesian Optimization. There are also population-based methods such as CMA-ES. In this paper, we present HBRKGA, a new population-based approach for hyperparameter optimization. HBRKGA is a hybrid approach that combines the Biased Random Key Genetic Algorithm with a Random Walk technique to search the hyperparameter space efficiently. Several computational experiments on eight different datasets were performed to assess the effectiveness of the proposed approach. Results showed that HBRKGA could find hyperparameter configurations that outperformed (in terms of predictive quality) the baseline methods in six out of eight datasets while showing a reasonable execution time.

Capturing Provenance from Deep Learning Applications Using Keras-Prov and Colab: a Practical Approach

Due to the exploratory nature of DNNs, DL specialists often need to modify the input dataset, change a filter when preprocessing input data, or fine-tune the models’ hyperparameters, while analyzing the evolution of the training. However, the specialist may lose track of what hyperparameter configurations have been used and tuned if these data are not properly registered. Thus, these configurations must be tracked and made available for the user’s analysis. One way of doing this is to use provenance data derivation traces to help the hyperparameter’s fine-tuning by providing a global data picture with clear dependencies. Current provenance solutions present provenance data disconnected from W3C PROV recommendation, which is difficult to reproduce and compare to other provenance data. To help with these challenges, we present Keras-Prov, an extension to the Keras deep learning library to collect provenance data compliant with PROV. To show the flexibility of Keras-Prov, we extend a previous Keras-Prov demonstration paper with larger experiments using GPUs with the help of Google Colab. Despite the challenges of running a DBMS with virtual environments, DL analysis with provenance has added trust and persistence in databases and PROV serializations. Experiments show Keras-Prov data analysis, during training execution, to support hyperparameter fine-tuning decisions, favoring the comparison, and reproducibility of such DL experiments. Keras-Prov is open source and can be downloaded from https://github.com/dbpina/keras-prov.

A Bayesian optimization hyperband-optimized incremental deep belief network for online battery behaviour modelling for a satellite simulator

Simulation tools play crucial roles in the stable implementation of space missions during satellite operations; they are typically utilized for satellite behaviour monitoring by comparing obtained telemetry values and predicted values according to pretrained prediction models. However, as telemetry data streams arrive in a chunk-by-chunk manner, a common practice is to retrain the employed simulation tool based on the newly arrived data, which results in the consumption of many computing resources and time lags. Therefore, an incremental learning algorithm is required to achieve accurate and fast satellite behaviour prediction. This paper proposes a Bayesian optimization hyperband-optimized incremental learning-based deep belief network (BOHB-ILDBN) to reproduce battery voltage behaviours, where the BOHB algorithm is utilized to obtain a group of optimal hyperparameter configurations to initialize a DBN model, the DBN model is incrementally updated by a fine-tuning process, and the variance difference between the actual and forecasted values is taken as the criterion for determining the completion of model training. Finally, the effectiveness and robustness of the model are verified on telemetry data obtained from an on-orbit sun-synchronous remote sensing satellite, the China–Brazil Earth Resources Satellite (CBERS-4A).

Artificial Neural Network Hyperparameters Optimization: A Survey

Machine-learning (ML) methods often utilized in applications like computer vision, recommendation systems, natural language processing (NLP), as well as user behavior analytics. Neural Networks (NNs) are one of the most es-sential ways to ML; the most challenging element of designing a NN is de-termining which hyperparameters to employ to generate the optimal model, in which hyperparameter optimization improves NN performance. This study includes a brief explanation regarding a few types of NN as well as some methods for hyperparameter optimization, as well as previous work results in enhancing ANN performance using optimization methods that aid research-ers and data analysts in developing better ML models via identifying the ap-propriate hyperparameter configurations.

Hyper-parameter optimization of multiple machine learning algorithms for molecular property prediction using hyperopt library

Due to outstanding performance in cheminformatics, machine learning algorithms have been increasingly used to mine molecular properties and biomedical big data. The performance of machine learning models is known to critically depend on the selection of the hyper-parameter configuration. However, many studies either explored the optimal hyper-parameters per the grid searching method or employed arbitrarily selected hyper-parameters, which can easily lead to achieving a suboptimal hyper-parameter configuration. In this study, Hyperopt Library embedding with the Bayesian optimization is employed to find optimal hyper-parameters for different machine learning algorithms. Six drug discovery datasets, including solubility, probe-likeness, hERG, Chagas disease, tuberculosis, and malaria, are used to compare different machine learning algorithms with ECFP6 fingerprints. This contribution aims to evaluate whether the Bernoulli Naïve Bayes, Logistic Linear Regression, AdaBoost Decision Tree, Random Forest, support vector machine, and deep neural networks algorithms with optimized hyper-parameters can offer any improvement in testing as compared with the Referenced Models assessed by an array of metrics including AUC, F1-score, Cohen’s kappa, Matthews correlation coefficient, recall, precision, and accuracy. Based on the rank normalized score approach, the Hyperopt Models achieve better or comparable performance on 33 out 36 models for different drug discovery datasets, showing significant improvement achieved by employing the Hyperopt library. The open-source code of all the 6 machine learning frameworks employed in the Hyperopt Python package is provided to make this approach accessible to more scientists, who are not familiar with writing code.

Graph-based mobility profiling

The decarbonization of the transport system requires a better understanding of human mobility behavior to optimally plan and evaluate sustainable transport options (such as Mobility as a Service). Current analysis frameworks often rely on specific datasets or data-specific assumptions and hence are difficult to generalize to other datasets or studies. In this work, we present a workflow to identify groups of users with similar mobility behavior that appear across several datasets. Our method does not depend on a specific clustering algorithm, is robust against the choice of hyperparameters, does not require specific labels in the dataset, and is not limited to specific types of tracking data. This allows the extraction of stable mobility profiles based on several small and inhomogeneous tracking data sets. Our method consists of the following main steps: Representing individual mobility using location-based graphs; extraction of graph-based mobility features; clustering using different hyperparameter configurations; group identification using statistical testing. The method is applied to six tracking datasets (Geolife, Green Class 1 + 2, yumuv and two Foursquare datasets) with a total of 1070 users that visit about 3′000’000 different locations with a total tracking duration of over 200′000 days. We can identify and interpret five mobility profiles that appear in all datasets and show how these profiles can be used to analyze longitudinal and cross-sectional tracking studies.

Hyperparameter optimization through context-based meta-reinforcement learning with task-aware representation

In this paper, we combine context-based Meta-Reinforcement Learning with task-aware representation to efficiently overcome data-inefficiency and limited generalization in the hyperparameter optimization problem. First, we propose a new context-based meta-RL model that disentangles task inference and control, which improves the meta-training efficiency and accelerates the learning process for unseen tasks. Second, the task properties are inferred on-line, which includes not only the dataset representation but also the task-solving experience, thus encouraging the agent to explore in a much smarter fashion. Third, we employ amortized meta-learning to meta-train the agent, which is simple and runs faster than the gradient-based meta-training method. Experimental results suggest that our method can search for the optimal hyperparameter configuration with limited computational cost in a reasonable time.

Understanding Bitcoin Price Prediction Trends under Various Hyperparameter Configurations

Since bitcoin has gained recognition as a valuable asset, researchers have begun to use machine learning to predict bitcoin price. However, because of the impractical cost of hyperparameter optimization, it is greatly challenging to make accurate predictions. In this paper, we analyze the prediction performance trends under various hyperparameter configurations to help them identify the optimal hyperparameter combination with little effort. We employ two datasets which have different time periods with the same bitcoin price to analyze the prediction performance based on the similarity between the data used for learning and future data. With them, we measure the loss rates between predicted values and real price by adjusting the values of three representative hyperparameters. Through the analysis, we show that distinct hyperparameter configurations are needed for a high prediction accuracy according to the similarity between the data used for learning and the future data. Based on the result, we propose a direction for the hyperparameter optimization of the bitcoin price prediction showing a high accuracy.

An auto-tuned hybrid deep learning approach for predicting fracture evolution

In this study, a novel auto-tuned hybrid deep learning approach composed of three base deep learning models, namely, long short-term memory, gated recurrent unit, and support vector regression, is developed to predict the fracture evolution process. The novelty of this framework lies in the auto-determined hyperparameter configurations for each base model based on the Bayesian optimization technique, which guarantees the fast and easy implementation in various practical applications. Moreover, the ensemble modeling technique auto consolidates the predictive capability of each base model to generate the final optimized hybrid model, which offers a better prediction of the overall fracture pattern evolution, as demonstrated by a case study. The comparison of the different prediction strategies exhibits that the direct prediction is a better option than the recursive prediction, in particular for a longer prediction distance. The proposed approach may be applied in various sequential data predictions by adopting the adaptive prediction scheme.

CrowdOptim: A Crowd-driven Neural Network Hyperparameter Optimization Approach to AI-based Smart Urban Sensing

AI-based smart urban sensing (ASUS) has emerged as a scalable and pervasive application paradigm in smart city planning and management that aims to automatically assess the physical status of the urban environments by leveraging AI techniques and massive urban sensing data. In this paper, we focus on a crowd-driven neural network (NN) hyperparameter optimization problem in ASUS applications. Our goal is to utilize the human intelligence from crowdsourcing systems to identify the optimal NN hyperparameter configuration for an ASUS model. Our work is motivated by the observation that the hyperparameters of current ASUS models are often manually configured by the AI specialists, which is known to be both error-prone and suboptimal. Two key technical challenges exist in solving our problem: i) it is challenging to effectively translate the highly complex NN hyperparameter optimization problem in AI to a simplified problem that can be solved by crowd workers without extensive AI expertise; ii) it is difficult to identify the optimal hyperparameter configuration in the large hyperparameter search space given the blackbox nature of the AI model. To address the above challenges, we develop CrowdOptim, a crowd-AI collaborative learning framework that integrates the techniques from crowdsourcing, hyperparameter optimization, and estimation theory to address the crowd-driven NN hyperparameter optimization problem in ASUS applications. The evaluation results from two real-world ASUS applications (i.e., smart city infrastructure monitoring (SCIM) and urban environment cleanliness assessment (UECA)) show that CrowdOptim consistently outperforms the state-of-the-art deep convolutional networks, crowd-AI, and hyperparameter optimization baselines in achieving the ASUS application objectives under various evaluation scenarios.

Fault Diagnosis of Rotating Machinery Based on Evolutionary Convolutional Neural Network

This paper proposes a fault diagnosis method for rotating machinery based on evolutionary convolutional neural network (ECNN). With the time-frequency images as the network input, with the help of the global optimization ability of the genetic algorithm, the structure of the convolutional neural network can evolve autonomously, and the adaptive configuration of the structural hyperparameters for the target task is realized. In this paper, the proposed method is verified by the measured signal of the planetary gearbox. The results show that the proposed method is helpful to obtain a convolutional neural network structure with better performance and achieve higher fault diagnosis accuracy.

MULTILINGUAL TEXT CLASSIFIER USING PRE-TRAINED UNIVERSAL SENTENCE ENCODER MODEL

Context. Online platforms and environments continue to generate ever-increasing content. The task of automating the moderation of user-generated content continues to be relevant. Of particular note are cases in which, for one reason or another, there is a very small amount of data to teach the classifier. To achieve results under such conditions, it is important to involve the classifier pre-trained models, which were trained on a large amount of data from a wide range. This paper deals with the use of the pre-trained multilingual Universal Sentence Encoder (USE) model as a component of the developed classifier and the affect of hyperparameters on the classification accuracy when learning on a small data amount (~ 0.05% of the dataset).
 Objective. The goal of this paper is the investigation of the pre-trained multilingual model and optimal hyperparameters influence for learning the text data classifier on the classification result.
 Method. To solve this problem, a relatively new approach to few-shot learning has recently been used – learning with a relatively small number of examples. Since text data is still the dominant way of transmitting information, the study of the possibilities of constructing a classifier of text data when learning from a small number of examples (~ 0.002–0.05% of the data set) is an actual problem.
 Results. It is shown that even with a small number of examples for learning (36 per class) due to the use of USE and optimal configuration in learning can achieve high accuracy of classification on English and Russian data, which is extremely important when it is impossible to collect your own large data set. The influence of the approach using USE and a set of different configurations of hyperparameters on the result of the text data classifier on the example of English and Russian data sets is evaluated.
 Conclusions. During the experiments, a significant degree of relevance of the correct selection of hyperparameters is shown. In particular, this paper considered the batch size, optimizer, number of learning epochs and the percentage of data from the set taken to train the classifier. In the process of experimentation, the optimal configuration of hyperparameters was selected, according to which 86.46% accuracy of classification on the Russian-language data set and 91.13% on the English-language data, respectively, can be achieved in ten seconds of training (training time can be significantly affected by technical means used).

Intelligent short term traffic forecasting using deep learning models with Bayesian contextual hyperband tuning

AbstractAn intelligent transport system (ITS) is fully valuable only if it can dynamically and aptly integrate all the latest cutting‐edge technologies. An ITS focuses on providing services like promptly offering real‐time road traffic information to interested parties, finding ways to reduce the average waiting time and offer secure and reliable services for commuters using past statistics. Short‐term traffic prediction is one such area in which the research community has focused in the past decade. Existing models developed for prediction has scope to improve in terms of accuracy and training time. There is a necessity to develop a best‐performing model that is computationally affluent to train with the optimal hyperparameter configuration as input which leads to improved performance. This article proposes a model that captures the traffic flow trend present in the past data to predict the flow for a future time interval. This model is an amalgamation of seasonal global trend (SGT) model and long short‐term memory (LSTM) model with attention mechanism. A novel hyperparameter tuning algorithm is also proposed which is based on multi‐armed bandit strategy with context, incorporating the right trade‐off between exploitation and exploration of the hyperparameter space using successive halving. Experimental results conducted proves that our forecast model in combination with the proposed hyperparameter tuning algorithm outperforms the existing models like SGT, LSTM models in terms of accuracy and time.

Receding-Horizon Prediction of Vehicle Velocity Profile Using Deterministic and Stochastic Deep Neural Network Models

The paper firstly proposes a deterministic deep feedforward neural network model aimed at predicting the city bus velocity profile over receding time horizon based on the following inputs: actual vehicle position, actual velocity or short-term history of vehicle velocities, time of day and day of week. A systematic analysis of the influence of different input subsets, history interval length and prediction horizon length is carried out to find an optimal configuration of NN model inputs and hyperparameters. Secondly, a stochastic version of neural network prediction model is proposed, which predicts expectations and standard deviations of velocity patterns over the receding time horizon. The stochastic model prediction accuracy is verified against the recorded test dataset features, as well as by comparing the predicted velocity expectation with the deterministic model prediction and correlating the predicted velocity standard deviation with deterministic model prediction uncertainty metrics. The verification results indicate that: (i) the deterministic model velocity prediction accuracy is characterized by the R2 score greater than 0.8 for the prediction horizon length of 10 s and remains to be solid (greater than 0.6) for the horizon lengths up to 25 s; (ii) the actual vehicle position and the velocity history are the most significant input features, where the optimal value of history interval length lies in the range from 30 to 50 s; (iii) the stochastic model have only slightly lower accuracy of predicting the velocity expectation along the receding horizon when compared to the deterministic model (the root mean square error is higher by 2.2%), and it outputs consistent standard deviation prediction.

Pavement Distress Recognition via Wavelet-Based Clustering of Smartphone Accelerometer Data

The ubiquity of smartphones has led to a variety of studies on how phone accelerometers can be used to assess road pavement quality. The majority of prior studies have emphasized supervised machine learning, assuming the ability to collect labeled data for model development. However, variances in vehicle dynamics and roadway quality, as well as the reliance on data labeling, limit the generalizability, scalability, and reproducibility of these approaches. Here, we propose an unsupervised learning framework that combines Pareto-optimized wavelet featurization and clustering. We first demonstrate the applicability of wavelets as features for dimensionally reduced accelerometer data. Wavelet featurization typically requires significant empirical tuning for optimal featurization. The presented framework automates tuning and selection of wavelet features based on subsequent clustering metrics. These metrics are related to inherent cluster characteristics such as intercluster variance and between-cluster variance. Rather than select a clustering method a priori, the proposed approach optimizes across a variety of clustering algorithms and hyperparameter configurations, again based on a set of clustering metrics. Experimental evaluation shows that the framework is able to detect road pavement distress and distinguish between classes of pavement defects, but that low-cost smartphone sensor data may not be as reliable in discriminating the more nuanced characteristics of pavement distress. This was most notable in the case of cracking, where the type and range of cracking severity caused undesirable cluster separation. The presented framework is general and cost-efficient and opens the way to further research on automatic pavement distress recognition from crowdsourced, low-cost data.

End-to-End Pipeline for Uncertainty Quantification and Remaining Useful Life Estimation: An Application on Aircraft Engines

Estimating the remaining useful life (RUL) of an asset lies at the heart of prognostics and health management (PHM) of many operations-critical industries such as aviation. Modern methods of RUL estimation adopt techniques from deep learning (DL). However, most of these contemporary techniques deliver only single-point estimates for the RUL without reporting on the confidence of the prediction. This practice usually provides overly confident predictions that can have severe consequences in operational disruptions or even safety. To address this issue, we propose a technique for uncertainty quantification (UQ) based on Bayesian deep learning (BDL). The hyperparameters of the framework are tuned using a novel bi-objective Bayesian optimization method with objectives the predictive performance and predictive uncertainty. The method also integrates the data pre-processing steps into the hyperparameter optimization (HPO) stage, models the RUL as a Weibull distribution, and returns the survival curves of the monitored assets to allow informed decision-making. We validate this method on the widely used C-MAPSS dataset against a single-objective HPO baseline that aggregates the two objectives through the harmonic mean (HM). We demonstrate the existence of trade-offs between the predictive performance and the predictive uncertainty and observe that the bi-objective HPO returns a larger number of hyperparameter configurations compared to the single-objective baseline. Furthermore, we see that with the proposed approach, it is possible to configure models for RUL estimation that exhibit better or comparable performance to the single-objective baseline when validated on the test sets.

Interpretable Assessment of ST-Segment Deviation in ECG Time Series.

Nowadays, even with all the tremendous advances in medicine and health protocols, cardiovascular diseases (CVD) continue to be one of the major causes of death. In the present work, we focus on a specific abnormality: ST-segment deviation, which occurs regularly in high-performance athletes and elderly people, serving as a myocardial infarction (MI) indicator. It is usually diagnosed manually by experts, through visual interpretation of the printed electrocardiography (ECG) signal. We propose a methodology to detect ST-segment deviation and quantify its scale up to 1 mV by extracting statistical, point-to-point beat characteristics and signal quality indexes (SQIs) from single-lead ECG. We do so by applying automated machine learning methods to find the best hyperparameter configuration for classification and regression models. For validation of our method, we use the ST-T database from Physionet; the results show that our method obtains 98.30% accuracy in the case of a multiclass problem and 99.87% accuracy in the case of binarization.