Given the significant advances in machine learning techniques on mobile devices, particularly in the domain of computer vision, in this work we quantitatively study the performance characteristics of 190 real-world vision transformers (ViTs) on mobile devices. Through a comparison with 102 real-world convolutional neural networks (CNNs), we provide insights into the factors that influence the latency of ViT architectures on mobile devices. Based on these insights, we develop a dataset including measured latencies of 1000 synthetic ViTs with representative building blocks and state-of-the-art architectures from two machine learning frameworks and six mobile platforms. Using this dataset, we show that inference latency of new ViTs can be predicted with sufficient accuracy for real-world applications.

In order to accurately identify the performance status of mobile devices and finely adjust the user experience, a real-time performance perception evaluation method based on TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) combined with entropy weighting method and time series model construction was studied. After collecting the performance characteristics of various mobile devices, the device performance profile was fitted by using PCA (principal component analysis) dimensionality reduction and feature engineering methods such as descriptive time series analysis. The ability of performance features and profiles to describe the real-time performance status of devices was understood and studied by applying the TOPSIS method and multi-level weighting processing. A time series model was constructed for the feature set under objective weighting, and multiple sensitivity (real-time, short-term, long-term) performance status perception results were provided to obtain real-time performance evaluation data and long-term stable performance prediction data. Finally, by configuring dynamic AB experiments and overlaying fine-grained power reduction strategies, the usability of the method was verified, and the accuracy of device performance status identification and prediction was compared with the performance of the profile features including dimensionality reduction time series modeling, TOPSIS method and entropy weighting method, subjective weighting, HMA method. The results show that accurate real-time performance perception results can greatly enhance business value, and this research has application effectiveness and certain forward-looking significance.

Feature selection is a crucial step in building machine learning models. This process is often achieved with accuracy as an objective, and can be cumbersome and computationally expensive for large-scale datasets. Several additional model performance characteristics such as fairness and robustness are of importance for model development. As regulations are driving the need for more trustworthy models, deployed models need to be corrected for model characteristics associated with responsible artificial intelligence. When feature selection is done with respect to one model performance characteristic (eg. accuracy), feature selection with secondary model performance characteristics (eg. fairness and robustness) as objectives would require going through the computationally expensive selection process from scratch. In this paper, we introduce the problem of feature \emph{reselection}, so that features can be selected with respect to secondary model performance characteristics efficiently even after a feature selection process has been done with respect to a primary objective. To address this problem, we propose REFRESH, a method to reselect features so that additional constraints that are desirable towards model performance can be achieved without having to train several new models. REFRESH's underlying algorithm is a novel technique using SHAP values and correlation analysis that can approximate for the predictions of a model without having to train these models. Empirical evaluations on three datasets, including a large-scale loan defaulting dataset show that REFRESH can help find alternate models with better model characteristics efficiently. We also discuss the need for reselection and REFRESH based on regulation desiderata.

Perception systems operate as a subcomponent of the general autonomy stack, and perception system designers often need to optimize performance characteristics while maintaining safety with respect to the overall closed-loop system. For this reason, it is useful to distill high-level safety requirements into component-level requirements on the perception system. In this work, we focus on efficiently determining sets of safe perception system performance characteristics given a black-box simulator of the fully-integrated, closed-loop system. We combine the advantages of common black-box estimation techniques such as Gaussian processes and threshold bandits to develop a new estimation method, which we call smoothing bandits. We demonstrate our method on a vision-based aircraft collision avoidance problem and show improvements in terms of both accuracy and efficiency over the Gaussian process and threshold bandit baselines.

This paper presents a multifunctional interdisciplinary framework that makes four scientific contributions towards the development of personalized ambient assisted living, with a specific focus to address the different and dynamic needs of the diverse aging population in the future of smart living environments. First, it presents a probabilistic reasoning-based mathematical approach to model all possible forms of user interactions for any activity arising from the user diversity of multiple users in such environments. Second, it presents a system that uses this approach with a machine learning method to model individual user profiles and user-specific user interactions for detecting the dynamic indoor location of each specific user. Third, to address the need to develop highly accurate indoor localization systems for increased trust, reliance, and seamless user acceptance, the framework introduces a novel methodology where two boosting approaches Gradient Boosting and the AdaBoost algorithm are integrated and used on a decision tree-based learning model to perform indoor localization. Fourth, the framework introduces two novel functionalities to provide semantic context to indoor localization in terms of detecting each user's floor-specific location as well as tracking whether a specific user was located inside or outside a given spatial region in a multi-floor-based indoor setting. These novel functionalities of the proposed framework were tested on a dataset of localization-related Big Data collected from 18 different users who navigated in 3 buildings consisting of 5 floors and 254 indoor spatial regions. The results show that this approach of indoor localization for personalized AAL that models each specific user always achieves higher accuracy as compared to the traditional approach of modeling an average user.

ABSTRACT A computer program that models an expert in a given domain is more likek to be accepted by experts in that domain, and by non-experts seeking its gavice. An explanation capability not only adds to the system's credibility, but also enables the non-expert user to learn from it. Furthermore, clear explanations allow an expert to check the system's "reasoning", possibly discovering the need for refinements and additions to the svstem's knowledge base. In a developing system, an explanation capability can be used as a debugging aid to verify that additions to the system are working as'hey should. The explanation facility in MYCIN is discussed as an illustration of how the various problems might be approached.