Felipe Farias

Felipe Costa Farias, Teresa Bernarda Ludermir, Carmelo Jose Albanez Bastos-Filho

The definition of a Neural Network architecture is one of the most critical and challenging tasks to perform. In this paper, we propose ParallelMLPs. ParallelMLPs is a procedure to enable the training of several independent Multilayer Perceptron Neural Networks with a different number of neurons and activation functions in parallel by exploring the principle of locality and parallelization capabilities of modern CPUs and GPUs. The core idea of this technique is to use a Modified Matrix Multiplication that replaces an ordinal matrix multiplication by two simple matrix operations that allow separate and independent paths for gradient flowing, which can be used in other scenarios. We have assessed our algorithm in simulated datasets varying the number of samples, features and batches using 10,000 different models. We achieved a training speedup from 1 to 4 orders of magnitude if compared to the sequential approach.

Felipe Costa Farias, Teresa Bernarda Ludermir, Carmelo José Albanez Bastos-Filho

The model selection procedure is usually a single-criterion decision making in which we select the model that maximizes a specific metric in a specific set, such as the Validation set performance. We claim this is very naive and can perform poor selections of over-fitted models due to the over-searching phenomenon, which over-estimates the performance on that specific set. Futhermore, real world data contains noise that should not be ignored by the model selection procedure and must be taken into account when performing model selection. Also, we have defined four theoretical optimality conditions that we can pursue to better select the models and analyze them by using a multi-criteria decision-making algorithm (TOPSIS) that considers proxies to the optimality conditions to select reasonable models.

Antonio Sa Barreto Neto, Felipe Farias, Marco Aurelio Tomaz Mialaret et al.

The increasing usage of smartphones in everyday tasks has been motivated many studies on energy consumption characterization aiming to improve smartphone devices' effectiveness and increase user usage time. In this scenario, it is essential to study mechanisms capable of characterizing user usage patterns, so smartphones' components can be adapted to promote the best user experience with lower energy consumption. The goal of this study is to build an energy consumption model based on user usage patterns aiming to provide the best accurate model to be used by application developers and automated optimization. To develop the energy consumption models, we established a method to identify the components with the most influence in the smartphone's energy consumption and identify the states of each influential device. Besides that, we established a method to prove the robustness of the models constructed using inaccurate hardware and a strategy to assess the accuracy of the model built. After training and testing each strategy to model the energy consumption based on the user's usage and perform the Nemenyi test, we demonstrated that it is possible to get a Mean Absolute Error of 158.57mW when the smartphone's average power is 1970.1mW. Some studies show that the leading smartphone's workload is the user. Based on this fact, we developed an automatic model building methodology that is capable of analyzing the user's usage data and build smart models that can estimate the smartphone's energy consumption based on the user's usage pattern. With the automatic model building methodology, we can adopt strategies to minimize the usage of components that drain the battery.

Felipe Farias, Teresa Ludermir, Carmelo Bastos-Filho

We propose a Similarity-Based Stratified Splitting (SBSS) technique, which uses both the output and input space information to split the data. The splits are generated using similarity functions among samples to place similar samples in different splits. This approach allows for a better representation of the data in the training phase. This strategy leads to a more realistic performance estimation when used in real-world applications. We evaluate our proposal in twenty-two benchmark datasets with classifiers such as Multi-Layer Perceptron, Support Vector Machine, Random Forest and K-Nearest Neighbors, and five similarity functions Cityblock, Chebyshev, Cosine, Correlation, and Euclidean. According to the Wilcoxon Sign-Rank test, our approach consistently outperformed ordinary stratified 10-fold cross-validation in 75\% of the assessed scenarios.