Benchmarking Deep Learning for Time Series: Challenges and Directions

Huang, Xinyuan; Fox, Geoffrey; Serebryakov, Sergey; Mohan, A. Krishna; Morkisz, Paweł; Dutta, Debojyoti

doi:10.1109/bigdata47090.2019.9005496

Cited by 13 publications

(6 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In future work, we will implement the super-resolution method confined to the bounding box areas. This will reduce the needed computational resources and allow us to use this method in real-time processing [37] and achieve better object recognition in this application. We will demonstrate this capability using modern streaming software environments [38].…”

Section: Resultsmentioning

confidence: 99%

Object Classifications by Image Super-Resolution Preprocessing for Convolutional Neural Networks

Na¹,

Fox²

2020

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

Blurred small objects produced by cropping, warping, or intrinsically so, are challenging to detect and classify. Therefore, much recent research is focused on feature extraction built on Faster R-CNN and follow-up systems. In particular, RPN, SPP, FPN, SSD, and DSSD are the layered feature extraction methods for multiple object detections and small objects. However, super-resolution methods, as explored here, can improve these image analyses working on before or after convolutional neural networks. Our methods are focused on building better image qualities into the original image components so that these feature extraction methods become more effective when applied later. Our super-resolution preprocessing resulted in better deep learning in the number of classified objects, especially for small objects when tested on the VOC2007, MSO, and COCO2017 datasets.

show abstract

Section: Resultsmentioning

confidence: 99%

Object Classifications by Image Super-Resolution Preprocessing for Convolutional Neural Networks

Na¹,

Fox²

2020

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

show abstract

“…Ruiz et al [48] and Javed et al [49] both presented comprehensive studies regarding the advances in and current state of the ML-based analysis of time-series data. Following the success of DL approaches, Huang et al [50] present a landscape of DL applications applied to time-series. Fawaz et al [18] studied the current state-of-the-art performance of DL algorithms for the classification of time-series data.…”

Section: Related Workmentioning

confidence: 99%

Benchmarking Transfer Learning Strategies in Time-Series Imaging: Recommendations for Analyzing Raw Sensor Data

2022

View full text Add to dashboard Cite

With the growing availability and complexity of time-series sequences, scalable and robust machine learning approaches are required that overcome the sampling challenge of quantitatively sufficient training data. Following the research trend towards the deep learning-based analysis of time-series encoded as images, this study proposes a time-series imaging workflow that overcomes the challenge of quantitatively limited sensor data across domains (i.e., medicine and engineering). After systematically identifying the three relevant dimensions that affect the performance of the deep learning-based analysis of visualized timeseries data, we performed a benchmarking evaluation with a total of 24 unique convolutional neural network models. Following a two-level transfer learning investigation, we reveal that fine-tuning the mid-level features results in the best classification performance. As a result, we present an optimized representation of the VGG16 network, which outperforms previous studies in the field. Our approach is accurate, robust, and manifests internal and external validity. By only using the raw time-series data, our model does not require manual feature engineering, being of high practical relevance. As the post-hoc analysis of our results reveals that our model allows automated extraction of meaningful features based on the trend of the underlying timeseries data, our study also adds to explainable artificial intelligence. Furthermore, our proposed workflow reduces the sequence length of the input data while preserving all information. Especially with the hurdle of long-term dependencies in sequential time-series data, we overcome related work's limitation of the vanishing gradients problem and contribute to the sequential learning theory in artificial intelligence.

show abstract

“…In this experiment, bandwidth requirement and flow duration prediction attained 90.67% and 91.33% accuracies, respectively. Huang et al [16] published a survey in 2019. It was based on deep learning use cases in the timeseries domain.…”

Section: The Advent Of 1d-cnn In Malware Detectionmentioning

confidence: 99%

Efficient Attack Detection in IoT Devices using Feature Engineering-Less Machine Learning

Khan¹,

Cotton²

2022

IJCSIT

View full text Add to dashboard Cite

Through the generalization of deep learning, the research community has addressed critical challenges in the network security domain, like malware identification and anomaly detection. However, they have yet to discuss deploying them on Internet of Things (IoT) devices for day-to-day operations. IoT devices are often limited in memory and processing power, rendering the compute-intensive deep learning environment unusable. This research proposes a way to overcome this barrier by bypassing feature engineering in the deep learning pipeline and using raw packet data as input. We introduce a feature- engineering-less machine learning (ML) process to perform malware detection on IoT devices. Our proposed model,” Feature- engineering-less-ML (FEL-ML),” is a lighter-weight detection algorithm that expends no extra computations on “engineered” features. It effectively accelerates the low-powered IoT edge. It is trained on unprocessed byte-streams of packets. Aside from providing better results, it is quicker than traditional feature-based methods. FEL-ML facilitates resource-sensitive network traffic security with the added benefit of eliminating the significant investment by subject matter experts in feature engineering.

show abstract

Benchmarking Deep Learning for Time Series: Challenges and Directions

Cited by 13 publications

References 28 publications

Object Classifications by Image Super-Resolution Preprocessing for Convolutional Neural Networks

Object Classifications by Image Super-Resolution Preprocessing for Convolutional Neural Networks

Benchmarking Transfer Learning Strategies in Time-Series Imaging: Recommendations for Analyzing Raw Sensor Data

Efficient Attack Detection in IoT Devices using Feature Engineering-Less Machine Learning

Contact Info

Product

Resources

About