Optimized Artificial Neural Network System to Select an Exploration Algorithm for Robots on Bi-dimensional Grids

Caballero, Liesle; Jojoa, Mario; Percybrooks, Winston S.

doi:10.1007/978-3-319-98998-3_2

Cited by 2 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The variable to be predicted in this application is defined as the maximum number of steps in which the robot reaches the target object for a given set of conditions. The experimental setup is described in [6] and [10]. The other three datasets were downloaded from the public UCI repository [11].…”

Section: Small Datasetsmentioning

confidence: 99%

“…Often, the hyper-parameters are determined by a random search inside a limited set of possibilities, picking the values that achieve the best performance. However, there is a optimization method proposed in [6] which calculates the number of layers and neurons per layer that maximizes the performance of the neural network for one or two layers and therefore, gets the best descriptive metrics of system behavior. This previous work detailed in [6] shows an improvement in predictor performance when Hill Climbing with Random Restart algorithm is applied to find the optimal ANN architecture comparing the results with a simple Hill Climbing algorithm, demonstrating that the results with the introduction of Random Restart are better.…”

Section: Hyper-parameter Optimizationmentioning

confidence: 99%

“…If a local optimum is found in the zero column of the search grid, it means that the structure of the neural network is single layer. Variations with respect to the rows of the zero column between one and twenty [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] determine the number of neurons in layer 1 when these structures are presented. For example, for the coordinate (20,0) shown in Fig.…”

Section: Hyper-parameter Optimizationmentioning

confidence: 99%

“…We test the performance of the optimization method using different datasets publicly available on the web, all with related existing literature, finding that the values of the evaluation metrics obtained when using our proposed method are better than the values reported in the Electronic supplementary material The online version of this article (https ://doi.org/10.1007/s4245 2-020-2060-5) contains supplementary material, which is available to authorized users. references [6][7][8][9]. The regression models predicted the following target variables: Performance of the Random Walk (RW) exploration algorithm, performance of the Random Walk Without Step Back (RW_WSB) exploration algorithm, performance of the Q_Learning exploration algorithm in autonomous robots, compressive strength of concrete, power consumed by a full-load plant and quality of red wine.…”

Section: Introductionmentioning

confidence: 99%

“…The machine learning technique used to design the predictor is a fully-connected, multilayer perceptron, artificial neural network (ANN) with one or two hidden layers, suitable to represent a non-linear model and connected by feed-forward links. Small network structures, i.e., few layers (one or two) and few of neurons per layer (< 20), are sufficient to solve regression problems applied to small data (400-10,000 samples) with a low number of features (4)(5)(6)(7)(8)(9)(10)(11).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Optimized neural networks in industrial data analysis

2020

Self Cite

View full text Add to dashboard Cite

The research work presented here tackles the problem of finding the optimum value for hyper-parameters, such as number of layers and number of neurons per layer, for a fully-connected Artificial Neural Network (ANN), particularly in regression problems. A proposed optimization strategy is tested on different datasets related to diverse industrial applications: (1) prediction of the performance of exploration algorithms for mobile robots, (2) prediction of the compressive strength of concrete, (3) prediction of energy output from a power plant; and (4) prediction of wine quality. Different evaluation metrics, such as Pearson correlation coefficient (R), Square Correlation Coefficient (R 2), Absolute Relative Error (RAE), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE), are used to determine the best performing prediction model when the hyper-parameter optimization algorithms are used. That result is compared to the performance of the best model previously reported in the literature on the same dataset in order to determine the gains achieved by using the hyper-parameter optimization strategy under test.

show abstract

Section: Small Datasetsmentioning

confidence: 99%

Section: Hyper-parameter Optimizationmentioning

confidence: 99%

Section: Hyper-parameter Optimizationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Optimized neural networks in industrial data analysis

2020

Self Cite

View full text Add to dashboard Cite

show abstract

Predicting the Energy Consumption of a Robot in an Exploration Task Using Optimized Neural Networks

Caballero¹,

Perafan

Rinaldy

et al. 2021

Electronics

Self Cite

View full text Add to dashboard Cite

This paper deals with the problem of determining a useful energy budget for a mobile robot in a given environment without having to carry out experimental measures for every possible exploration task. The proposed solution uses machine learning models trained on a subset of possible exploration tasks but able to make predictions on untested scenarios. Additionally, the proposed model does not use any kinematic or dynamic models of the robot, which are not always available. The method is based on a neural network with hyperparameter optimization to improve performance. Tabu List optimization strategy is used to determine the hyperparameter values (number of layers and number of neurons per layer) that minimize the percentage relative absolute error (%RAE) while maximize the Pearson correlation coefficient (R) between predicted data and actual data measured under a number of experimental conditions. Once the optimized artificial neural network is trained, it can be used to predict the performance of an exploration algorithm on arbitrary variations of a grid map scenario. Based on such prediction, it is possible to know the energy needed for the robot to complete the exploration task. A total of 128 tests were carried out using a robot executing two exploration algorithms in a grid map with the objective of locating a target whose location is not known a priori by the robot. The experimental energy consumption was measured and compared with the prediction of our model. A success rate of 96.093% was obtained, measured as the percentage of tests where the energy budget suggested by the model was enough to actually carry out the task when compared to the actual energy consumed in the test, suggesting that the proposed model could be useful for energy budgeting in actual mobile robot applications.

show abstract

Optimized Artificial Neural Network System to Select an Exploration Algorithm for Robots on Bi-dimensional Grids

Cited by 2 publications

References 12 publications

Optimized neural networks in industrial data analysis

Optimized neural networks in industrial data analysis

Predicting the Energy Consumption of a Robot in an Exploration Task Using Optimized Neural Networks

Contact Info

Product

Resources

About