Compressed Learning for Image Classification: A Deep Neural Network Approach

Zisselman, Ev; Adler, Amir; Elad, Michael

doi:10.1016/bs.hna.2018.08.002

Cited by 39 publications

(30 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The concept of dimensionality reduction in PCA pitches its use in facial recognition, computer vision and image compression. It has also wide spectrum of applications in pattern identification of high dimensional data pertaining to the field of finance, datamining, bio-informatics and psychology [30][31][32][33].…”

Section: Principal Component Analysismentioning

confidence: 99%

Early Detection of Diabetic Retinopathy Using PCA-Firefly Based Deep Learning Model

et al. 2020

View full text Add to dashboard Cite

Diabetic Retinopathy is a major cause of vision loss and blindness affecting millions of people across the globe. Although there are established screening methods - fluorescein angiography and optical coherence tomography for detection of the disease but in majority of the cases, the patients remain ignorant and fail to undertake such tests at an appropriate time. The early detection of the disease plays an extremely important role in preventing vision loss which is the consequence of diabetes mellitus remaining untreated among patients for a prolonged time period. Various machine learning and deep learning approaches have been implemented on diabetic retinopathy dataset for classification and prediction of the disease but majority of them have neglected the aspect of data pre-processing and dimensionality reduction, leading to biased results. The dataset used in the present study is a diabetes retinopathy dataset collected from the UCI machine learning repository. At its inceptions, the raw dataset is normalized using the Standardscalar technique and then Principal Component Analysis (PCA) is used to extract the most significant features in the dataset. Further, Firefly algorithm is implemented for dimensionality reduction. This reduced dataset is fed into a Deep Neural Network Model for classification. The results generated from the model is evaluated against the prevalent machine learning models and the results justify the superiority of the proposed model in terms of Accuracy, Precision, Recall, Sensitivity and Specificity.

show abstract

Section: Principal Component Analysismentioning

confidence: 99%

Early Detection of Diabetic Retinopathy Using PCA-Firefly Based Deep Learning Model

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Furthermore, in PCA, the dimensional reduction concept is utilised for image compression, computer vision, and facial recognition. It also presents a wide range of applications in determining the high-dimensional record patterns related to data mining, finance, psychology, and bioinformatics [27][28][29][30].…”

Section: Principle Component Analysis (Pca)mentioning

confidence: 99%

“…The feature values in the PD dataset have different ranges, which leads to noise in classification performance. Therefore, normalisation is performed on the dataset [30] to set the data in a uniform range of [0,1]. Normalisation is calculated using the following Equation:…”

Section: Normalizingmentioning

confidence: 99%

“…Accuracy is a misleading evaluation metric for the majority class and seldom predicts the parameters belonging to the minority class. For the classification problems involving unbalanced class distribution, we seek highly comprehensive evaluation by considering various aspects comprising the measurement of the classifier ability to attain a balance between two classes and considering both classes to be similar [30]. In our experiments, the area under the curve (AUC) and geometric imply (G-mean) are used because they exhibit robustness for unbalanced data distribution within the data.…”

Section: Evaluation Metrics Of Hd Performancementioning

confidence: 99%

See 1 more Smart Citation

Hybrid Feature Selection Framework for the Parkinson Imbalanced Dataset Prediction Problem

et al. 2021

View full text Add to dashboard Cite

Background and Objectives: Recently, many studies have focused on the early detection of Parkinson’s disease (PD). This disease belongs to a group of neurological problems that immediately affect brain cells and influence the movement, hearing, and various cognitive functions. Medical data sets are often not equally distributed in their classes and this gives a bias in the classification of patients. We performed a Hybrid feature selection framework that can deal with imbalanced datasets like PD. Use the SOMTE algorithm to deal with unbalanced datasets. Removing the contradiction from the features in the dataset and decrease the processing time by using Recursive Feature Elimination (RFE), and Principle Component Analysis (PCA). Materials and Methods: PD acoustic datasets and the characteristics of control subjects were used to construct classification models such as Bagging, K-nearest neighbour (KNN), multilayer perceptron, and the support vector machine (SVM). In the prepressing stage, the synthetic minority over-sampling technique (SMOTE) with two-feature selection RFE and PCA were used. The PD dataset comprises a large difference between the numbers of the infected and uninfected patients, which causes the classification bias problem. Therefore, SMOTE was used to resolve this problem. Results: For model evaluation, the train–test split technique was used for the experiment. All the models were Grid-search tuned, the evaluation results of the SVM model showed the highest accuracy of 98.2%, and the KNN model exhibited the highest specificity of 99%. Conclusions: the proposed method is compared with the current modern methods of detecting Parkinson’s disease and other methods for medical diseases, it was noted that our developed system could treat data bias and reach a high prediction of PD and this can be beneficial for health organizations to properly prioritize assets.

show abstract

“…There have been studies that used sensing to conserve power and communication and reduce latency. Compressed Learning [22] makes observations in a way that reduces the number of samples required for detection, and there has been research on conserving communication by computing the earlier parts of the deep neural network (DNN) model at the sensors then transmitting the results from the intermediate layers, where there are fewer nodes, to the server to compute the remainder of the model [23]. In particular, when using a user's smartphone for crowd sensing, conserving power and volume of communication are major issues, so application of these methods should reduce the burden on users and make it easier for them to work with our method.…”

Section: Related Researchmentioning

confidence: 99%

Barrier Detection Using Sensor Data from Multiple Transportation Modes

Kurauchi¹,

Abe²,

Matsuda³

et al. 2020

Journal of Information Processing

View full text Add to dashboard Cite

Providing routes that are passable allowing movement by transportation modes such as wheelchairs or strollers requires accessibility information including details about the type and location of barriers. Earlier research detected the types of barriers using sensor data from people in wheelchairs and able-bodied people, but the amount and range of travel of people in wheelchairs is limited. Also, small barriers are not noticed by able-bodied persons since they tend to be easily ignored during movement. In our research, the goal was to detect barrier details using sensor data from various transportation modes. However, there are issues with the selection of barrier detection models for each mode and with the cost of collecting data. To overcome this model-selection problem, we propose a model that detects transportation modes and barriers in two stages. We also propose a method for reducing the cost of collecting data, with which we prepare a course with a smooth surface, collect data, and simulate rough surfaces by adding noise. We conducted three experiments to verify the effectiveness of the proposed method. The proposed method achieved an accuracy of 91.5% in detecting six transportation modes, and showed that adding noise increased accuracy by 3.7 percentage points on rough surfaces. When detecting eight types of barriers, our method achieved an accuracy of 87.7% for walking with a stroller, and showed that adding noise increased accuracy by 6.8 percentage points on rough surfaces. Therefore, the proposed method is effective in detecting barrier details using multiple transportation modes.

show abstract

Compressed Learning for Image Classification: A Deep Neural Network Approach

Cited by 39 publications

References 16 publications

Early Detection of Diabetic Retinopathy Using PCA-Firefly Based Deep Learning Model

Early Detection of Diabetic Retinopathy Using PCA-Firefly Based Deep Learning Model

Hybrid Feature Selection Framework for the Parkinson Imbalanced Dataset Prediction Problem

Barrier Detection Using Sensor Data from Multiple Transportation Modes

Contact Info

Product

Resources

About