An Interpretable Machine Learning Model Enhanced Integrated CPU-GPU DVFS Governor

2021 ◽  
Vol 20 (6) ◽  
pp. 1-28
Author(s):  
Jurn-Gyu Park ◽  
Nikil Dutt ◽  
Sung-Soo Lim

Modern heterogeneous CPU-GPU-based mobile architectures, which execute intensive mobile gaming/graphics applications, use software governors to achieve high performance with energy-efficiency. However, existing governors typically utilize simple statistical or heuristic models, assuming linear relationships using a small unbalanced dataset of mobile games; and the limitations result in high prediction errors for dynamic and diverse gaming workloads on heterogeneous platforms. To overcome these limitations, we propose an interpretable machine learning (ML) model enhanced integrated CPU-GPU governor: (1) It builds tree-based piecewise linear models (i.e., model trees) offline considering both high accuracy (low error) and interpretable ML models based on mathematical formulas using a simulatability operation counts quantitative metric. And then (2) it deploys the selected models for online estimation into an integrated CPU-GPU Dynamic Voltage Frequency Scaling governor. Our experiments on a test set of 20 mobile games exhibiting diverse characteristics show that our governor achieved significant energy efficiency gains of over 10% (up to 38%) improvements on average in energy-per-frame with a surprising-but-modest 3% improvement in Frames-per-Second performance, compared to a typical state-of-the-art governor that employs simple linear regression models.

Author(s):  
Mark Endrei ◽  
Chao Jin ◽  
Minh Ngoc Dinh ◽  
David Abramson ◽  
Heidi Poxon ◽  
...  

Rising power costs and constraints are driving a growing focus on the energy efficiency of high performance computing systems. The unique characteristics of a particular system and workload and their effect on performance and energy efficiency are typically difficult for application users to assess and to control. Settings for optimum performance and energy efficiency can also diverge, so we need to identify trade-off options that guide a suitable balance between energy use and performance. We present statistical and machine learning models that only require a small number of runs to make accurate Pareto-optimal trade-off predictions using parameters that users can control. We study model training and validation using several parallel kernels and more complex workloads, including Algebraic Multigrid (AMG), Large-scale Atomic Molecular Massively Parallel Simulator, and Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics. We demonstrate that we can train the models using as few as 12 runs, with prediction error of less than 10%. Our AMG results identify trade-off options that provide up to 45% improvement in energy efficiency for around 10% performance loss. We reduce the sample measurement time required for AMG by 90%, from 13 h to 74 min.


2020 ◽  
Author(s):  
Lucas M. Thimoteo ◽  
Marley M. Vellasco ◽  
Jorge M. do Amaral ◽  
Karla Figueiredo ◽  
Cátia Lie Yokoyama ◽  
...  

This work proposes an interpretable machine learning approach to diagnosesuspected COVID-19 cases based on clinical variables. Results obtained for the proposed models have F-2 measure superior to 0.80 and accuracy superior to 0.85. Interpretation of the linear model feature importance brought insights about the most relevant features. Shapley Additive Explanations were used in the non-linear models. They were able to show the difference between positive and negative patients as well as offer a global interpretability sense of the models.


2009 ◽  
Vol 12 (3) ◽  
pp. 262-274 ◽  
Author(s):  
Nahm-Chung Jung ◽  
Ioana Popescu ◽  
Peter Kelderman ◽  
Dimitri P. Solomatine ◽  
Roland K. Price

A promising new approach for eco-environmental modelling, such as algal growth prediction, is the data-driven modeling using machine learning techniques: an artificial neural network (ANN) being a typical method. Another method growing in popularity, based on the M5 model tree (MT) algorithm, is the use of piecewise linear regression models at the leaf nodes of the tree. M5 MTs using partial least-squares regression (PLSR) proposed in this paper were tested on a particular dataset and then compared to M5 MTs, MLF- and RBF-ANN and k nearest neighbours (kNN). With the dataset partitioned to periods of algal growth and no growth, M5 MTs using PLSR showed better results for algal growth prediction in the reservoir than using the annual dataset and other algorithms. This gives the idea that the M5-PLSR MTs, in spite of the lack of data, more effectively seeks latent vectors between the closely correlated multivariate dataset partitioned using clustering techniques. M5-PLSR MTs is a promising approach when there is a shortage of data required to build a more transparent learning process model, and a combination with clustering is recommended.


2021 ◽  
Vol 7 ◽  
pp. e813
Author(s):  
Anandan Chinnalagu ◽  
Ashok Kumar Durairaj

Customer satisfaction and their positive sentiments are some of the various goals for successful companies. However, analyzing customer reviews to predict accurate sentiments have been proven to be challenging and time-consuming due to high volumes of collected data from various sources. Several researchers approach this with algorithms, methods, and models. These include machine learning and deep learning (DL) methods, unigram and skip-gram based algorithms, as well as the Artificial Neural Network (ANN) and bag-of-word (BOW) regression model. Studies and research have revealed incoherence in polarity, model overfitting and performance issues, as well as high cost in data processing. This experiment was conducted to solve these revealing issues, by building a high performance yet cost-effective model for predicting accurate sentiments from large datasets containing customer reviews. This model uses the fastText library from Facebook’s AI research (FAIR) Lab, as well as the traditional Linear Support Vector Machine (LSVM) to classify text and word embedding. Comparisons of this model were also done with the author’s a custom multi-layer Sentiment Analysis (SA) Bi-directional Long Short-Term Memory (SA-BLSTM) model. The proposed fastText model, based on results, obtains a higher accuracy of 90.71% as well as 20% in performance compared to LSVM and SA-BLSTM models.


Electronics ◽  
2021 ◽  
Vol 10 (17) ◽  
pp. 2113
Author(s):  
Qasem Abu Al-Haija ◽  
Abdulaziz A. Alsulami

The Bitcoin cryptocurrency is a worldwide prevalent virtualized digital currency conceptualized in 2008 as a distributed transactions system. Bitcoin transactions make use of peer-to-peer network nodes without a third-party intermediary, and the transactions can be verified by the node. Although Bitcoin networks have exhibited high efficiency in the financial transaction systems, their payment transactions are vulnerable to several ransomware attacks. For that reason, investigators have been working on developing ransomware payment identification techniques for bitcoin transactions’ networks to prevent such harmful cyberattacks. In this paper, we propose a high performance Bitcoin transaction predictive system that investigates the Bitcoin payment transactions to learn data patterns that can recognize and classify ransomware payments for heterogeneous bitcoin networks. Specifically, our system makes use of two supervised machine learning methods to learn the distinguishing patterns in Bitcoin payment transactions, namely, shallow neural networks (SNN) and optimizable decision trees (ODT). To validate the effectiveness of our solution approach, we evaluate our machine learning based predictive models on a recent Bitcoin transactions dataset in terms of classification accuracy as a key performance indicator and other key evaluation metrics such as the confusion matrix, positive predictive value, true positive rate, and the corresponding prediction errors. As a result, our superlative experimental result was registered to the model-based decision trees scoring 99.9% and 99.4% classification detection (two-class classifier) and accuracy (multiclass classifier), respectively. Hence, the obtained model accuracy results are superior as they surpassed many state-of-the-art models developed to identify ransomware payments in bitcoin transactions.


2019 ◽  
Author(s):  
Vinícius Klôh ◽  
Matheus Gritz ◽  
Bruno Schulze ◽  
Mariza Ferro

Performance and energy efficiency are now critical concerns in high performance scientific computing. It is expected that requirements of the scientific problem should guide the orchestration of different techniques of energy saving, in order to improve the balance between energy consumption and application performance. To enable this balance, we propose the development of an autonomous framework to make this orchestration and present the ongoing research to this development, more specifically, focusing in the characterization of the scientific applications and the performance modeling tasks using Machine Learning.


Sign in / Sign up

Export Citation Format

Share Document