On Parameter Tuning in Meta-Learning for Computer Vision

Mohammadi, Farid Ghareh; Arabnia, Hamid R.; Amini, M. Hadi

doi:10.1109/csci49370.2019.00060

Cited by 10 publications

(9 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deploying Machine Learning (ML) for real-world problems causes a number of challenges, e.g., selecting the proper model among a set of candidate models, hyper-parameter tuning, and selecting the dataset's features to feed to ML models. An ML model's performance depends on such initial design decisions, which can be confusing to new users who want to choose the suitable model [1]. These decisions need to be made based on some selection criteria, and are usually based on the model's obtained quality (performance indicator) [2].…”

Section: A Motivationmentioning

confidence: 99%

“…These decisions need to be made based on some selection criteria, and are usually based on the model's obtained quality (performance indicator) [2]. According to the principle of Occam's razor, a model should not be too simple nor too complex so that it can be efficient (in regards to computations) and at the same time capture data patterns, without overfitting [1], [3]. Hyper-parameter tuning is choosing the parameters' values that have a more considerable impact on the ML model's final performance (e.g., accuracy and run-time).…”

Section: A Motivationmentioning

confidence: 99%

See 1 more Smart Citation

HyP-ABC: A Novel Automated Hyper-Parameter Tuning Algorithm Using Evolutionary Optimization

Zahedi¹,

Mohammadi²,

Amini³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Section: A Motivationmentioning

confidence: 99%

Section: A Motivationmentioning

confidence: 99%

HyP-ABC: A Novel Automated Hyper-Parameter Tuning Algorithm Using Evolutionary Optimization

Zahedi¹,

Mohammadi²,

Amini³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…This approach took advantage of one of the important data analytic promises called zero-shot learning (ZSL) [41,7] . There are many applications of ZSL in computer vision [42] and motion detection in healthcare application [43] Furthermore, researchers in [44] proposed a solution based on ZSL called Zero-Virus providing a deep understanding for an intelligent transportation system to generate the best rout for drivers. Zero-virus does not need any vehicle-tracklets annotation, thus it is the most volatile real-world traffic scheme.…”

Section: Machine Learning Process and Challengesmentioning

confidence: 99%

Data Analytics for Smart cities: Challenges and Promises

Mohammadi¹,

Shenavarmasouleh²,

Amini³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

The explosion of advancements in artificial intelligence, sensor technologies, and wireless communication activates ubiquitous sensing through distributed sensors. These sensors are various domains of networks that lead us to smart systems in healthcare, transportation, environment, and other relevant branches/networks. Having collaborative interaction among the smart systems connects end-user devices to each other which enables achieving a new integrated entity called Smart Cities. The goal of this study is to provide a comprehensive survey of data analytics in smart cities. In this paper, we aim to focus on one of the smart cities important branches, namely Smart Mobility, and its positive ample impact on the smart cities decision-making process. Intelligent decision-making systems in smart mobility offer many advantages such as saving energy, relaying city traffic, and more importantly, reducing air pollution by offering real-time useful information and imperative knowledge. Making a decision in smart cities in time is challenging due to various and high dimensional factors and parameters, which are not frequently collected. In this paper, we first address current challenges in smart cities and provide an overview of potential solutions to these challenges. Then, we offer a framework of these solutions, called universal smart cities decision making, with three main sections of data capturing, data analysis, and decision making to optimize the smart mobility within smart cities. With this framework, we elaborate on fundamental concepts of big data, machine learning, and deep leaning algorithms that have been applied to smart cities and discuss the role of these algorithms in decision making for smart mobility in smart cities.

show abstract

“…decisions, which can be confusing to new users who desire to choose the most appropriate model [1]. These decisions are usually made based on the model's obtained quality or in other words Performance Indicator (PI) [2].…”

mentioning

confidence: 99%

“…These decisions are usually made based on the model's obtained quality or in other words Performance Indicator (PI) [2]. According to the principle of Occam's razor, a model should not be too simple nor too complex so that it can be efficient and can also capture data patterns without overfitting [1], [3]. Hyper-parameter tuning chooses the values that have a more considerable impact on the ML model's performance.…”

mentioning

confidence: 99%

HyP-ABC: A Novel Automated Hyper-Parameter Tuning Algorithm Using Evolutionary Optimization

Zahedi¹,

Mohammadi²,

Amini³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

<p>Machine learning techniques lend themselves as promising decision-making and analytic tools in a wide range of applications. Different ML algorithms have various hyper-parameters. In order to tailor an ML model towards a specific application working at its best, its hyper-parameters should be tuned. Tuning the hyper-parameters directly affects the performance. However, for large-scale search spaces, efficiently exploring the ample number of combinations of hyper-parameters is computationally expensive. Many of the automated hyper-parameter tuning techniques suffer from low convergence rates and high experimental time complexities. In this paper, we propose HyP-ABC, an automatic innovative hybrid hyper-parameter optimization algorithm using the modified artificial bee colony approach, to measure the classification accuracy of three ML algorithms: random forest, extreme gradient boosting, and support vector machine. In order to ensure the robustness of the proposed method, the algorithm takes a wide range of feasible hyper-parameter values and is tested using a real-world educational dataset. Experimental results show that HyP-ABC is competitive with state-of-the-art techniques. Also, it has fewer hyper-parameters to be tuned than other population-based algorithms, making it worthwhile for real-world HPO problems.</p>

show abstract

On Parameter Tuning in Meta-Learning for Computer Vision

Cited by 10 publications

References 35 publications

HyP-ABC: A Novel Automated Hyper-Parameter Tuning Algorithm Using Evolutionary Optimization

HyP-ABC: A Novel Automated Hyper-Parameter Tuning Algorithm Using Evolutionary Optimization

Data Analytics for Smart cities: Challenges and Promises

HyP-ABC: A Novel Automated Hyper-Parameter Tuning Algorithm Using Evolutionary Optimization

Contact Info

Product

Resources

About