Multi-objective evolution of artificial neural networks in multi-class medical diagnosis problems with class imbalance

Shenfield, Alex; Rostami, Shahin

doi:10.1109/cibcb.2017.8058553

Cited by 17 publications

(7 citation statements)

References 42 publications

(46 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep Convolutional Neural Network (CNN) was used extensively for ECG analysis [45,46,47,48,49]. Deep Artificial Neural Network (ANN) have be used for multiple heath care applications [50,51,52].…”

Section: Discussionmentioning

confidence: 99%

Automated detection of atrial fibrillation using long short-term memory network with RR interval signals

Faust

Shenfield

Kareem

et al. 2018

Computers in Biology and Medicine

Self Cite

245

131

View full text Add to dashboard Cite

Atrial Fibrillation (AF), either permanent or intermittent (paroxysnal AF), increases the risk of cardioembolic stroke. Accurate diagnosis of AF is obligatory for initiation of effective treatment to prevent stroke. Long term cardiac monitoring improves the likelihood of diagnosing paroxysmal AF. We used a deep learning system to detect AF beats in Heart Rate (HR) signals. The data was partitioned with a sliding window of 100 beats. The resulting signal blocks were directly fed into a deep Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM). The system was validated and tested with data from the MIT-BIH Atrial Fibrillation Database. It achieved 98.51% accuracy with 10-fold cross-validation (20 subjects) and 99.77% with blindfold validation (3 subjects). The proposed system structure is straight forward, because there is no need for information reduction through feature extraction. All the complexity resides in the deep learning system, which gets the entire information from a signal block. This setup leads to the robust performance for unknown data, as measured with the blind fold validation. The proposed Computer-Aided Diagnosis (CAD) system can be used for long-term monitoring of the human heart. To the best of our knowledge, the proposed system is the first to incorporate deep learning for AF beat detection.

show abstract

Section: Discussionmentioning

confidence: 99%

Automated detection of atrial fibrillation using long short-term memory network with RR interval signals

Faust

Shenfield

Kareem

et al. 2018

Computers in Biology and Medicine

Self Cite

245

131

View full text Add to dashboard Cite

show abstract

“…The concept of chromosome spectrum was introduced and used, together with a novelty map-population to achieve this diversity. Direct encoding of Markov networks was facilitated by genetic programing in [22], and in [23,24], the structures, weights, and biases of a network were optimized employing a multi-objective evolution strategy with a preference articulation.…”

Section: Artificial Neural Networkmentioning

confidence: 99%

“…In the above equations, α and β are known, and the other values can be set to arbitrary values. Solving Equations (23) and (24) provided the corresponding values α and β.…”

Section: Feature Conversionmentioning

confidence: 99%

3D object recognition with a linear time‐varying system of overlay layers

Nasrabadi

Safabakhsh

2021

IET Computer Vision

View full text Add to dashboard Cite

Object recognition is a challenging task in computer vision with numerous applications. The challenge is in selecting appropriate robust features with tolerable computing costs. Feature learning attempts to solve the feature extraction problem through a learning process using various samples of the objects. This research proposes a two-stage optimization framework to identify the structure of a first-order linear non-homogeneous difference equation which is a linear time-varying system of overlay layers (LtvoL) that construct an image. The first stage consists of the determination of a finite set of impulses, called overlay layers, by the application of a genetic algorithm. The second stage defines the coefficients of the corresponding difference equation derived from L 2 regularization. Classification of the test images is possible by a novel process exclusively designed for this model. Experiments on the Washington RGB-D dataset and ETH-80 show promising results which are comparable to those of state-of-the-art methods for RGB-D-based object recognition.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

“…A novel multi-objective evolutionary algorithm (ENORA) was created to search for and select the optimal feature subset in the context of a multi-class classification problem [19]. Shenfield and Rostami [20] apply an evolutionary algorithm that optimizes neural network weights, biases, and structures to simultaneously optimize both overall and individual class accuracy. In RapidMiner [21], an evolutionary framework is proposed where the user may manually design the evolutionary algorithm using drag and drop features.…”

Section: Related Workmentioning

confidence: 99%

“…Unlike common multi-objective optimization approaches that solely use metaheuristics [20], [23], [25], the default approach employed by Autotune is a novel hybrid strategy that combines the global search emphasis of metaheuristic [38] with lesser known, but efficient, direct local search methods [39]. The hybrid search strategy begins by creating a Latin Hypercube Sampling (LHS) of the search space.…”

Section: B Multi-objective Optimization Approachmentioning

confidence: 99%

Constrained Multi-Objective Optimization for Automated Machine Learning

Gardner

Golovidov

Griffin

et al. 2019

2019 IEEE International Conference on Data Science and Advanced Analytics (DSAA)

View full text Add to dashboard Cite

Automated machine learning has gained a lot of attention recently. Building and selecting the right machine learning models is often a multi-objective optimization problem. General purpose machine learning software that simultaneously supports multiple objectives and constraints is scant, though the potential benefits are great. In this work, we present a framework called Autotune that effectively handles multiple objectives and constraints that arise in machine learning problems. Autotune is built on a suite of derivative-free optimization methods, and utilizes multi-level parallelism in a distributed computing environment for automatically training, scoring, and selecting good models. Incorporation of multiple objectives and constraints in the model exploration and selection process provides the flexibility needed to satisfy trade-offs necessary in practical machine learning applications. Experimental results from standard multi-objective optimization benchmark problems show that Autotune is very efficient in capturing Pareto fronts. These benchmark results also show how adding constraints can guide the search to more promising regions of the solution space, ultimately producing more desirable Pareto fronts. Results from two real-world case studies demonstrate the effectiveness of the constrained multiobjective optimization capability offered by Autotune.

show abstract

Multi-objective evolution of artificial neural networks in multi-class medical diagnosis problems with class imbalance

Cited by 17 publications

References 42 publications

Automated detection of atrial fibrillation using long short-term memory network with RR interval signals

Automated detection of atrial fibrillation using long short-term memory network with RR interval signals

3D object recognition with a linear time‐varying system of overlay layers

Constrained Multi-Objective Optimization for Automated Machine Learning

Contact Info

Product

Resources

About