Classification Techniques in Machine Learning: Applications and Issues

Soofi, Aized Amin; Awan, Arshad

doi:10.6000/1927-5129.2017.13.76

Cited by 231 publications

(81 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…k-NN algorithms are widely used in applications such as face recognition (Kasemsumran et al 2016), traffic forecasting (Zhang et al 2013) and speaking recognition (Rizwan and Anderson 2014) achieving high performance when large training datasets are available. It is robust to noisy data and easy to visualize but it often requires large memory allocation (Soofi and Awan 2017).…”

Section: Machine Learning Algorithmsmentioning

confidence: 99%

“…The margin corresponds to the shortest distance between the closest points to the boundary hyperplane function. SVM is able to deal with a large variety of classification problems including high dimensional and non-linear problems (Soofi and Awan 2017). Although SVM is very powerful, it is difficult to visualise (Karamizadeh et al 2014) and require accurate often a priori selection of a number of parameters within the kernel function.…”

Section: Machine Learning Algorithmsmentioning

confidence: 99%

See 1 more Smart Citation

Use of acoustic emission in combination with machine learning: monitoring of gas–liquid mixing in stirred tanks

et al. 2020

View full text Add to dashboard Cite

Operations involving gas-liquid agitated vessels are common in the biochemical and chemical industry; ensuring good contact between the two phases is essential to process performance. In this work, a methodology to compute acoustic emission data, recorded using a piezoelectric sensor, to evaluate the gas-liquid mixing regime within gas-liquid and gas-solid-liquid mixtures was developed. The system was a 3L stirred tank equipped with a Rushton Turbine and a ring sparger. Whilst moving up through the vessel, gas bubbles collapse, break or coalesce generating sound waves transmitted through the wall to the acoustic transmitter. The system was operated in different flow regimes (non-gassed condition, loaded and complete dispersion) achieved by varying impeller speed and gas flow rate, with the objective to feed machine learning algorithms with the acoustic spectrum to univocally identify the different conditions. The developed method allowed to successfully recognise the operating regime with an accuracy higher than 90% both in absence and presence of suspended particles.

show abstract

Section: Machine Learning Algorithmsmentioning

confidence: 99%

Section: Machine Learning Algorithmsmentioning

confidence: 99%

Use of acoustic emission in combination with machine learning: monitoring of gas–liquid mixing in stirred tanks

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Classification technique is widely used to classify the data on the basis of its features into different classes. There are various machine learning classification techniques . In this paper, three classification algorithms have been used as discussed in Section : Multinomial logistic regression Multinomial logistic regression is also known as polytomous.…”

Section: Methodsmentioning

confidence: 99%

ORuML: Optimized Routing in wireless networks using Machine Learning

Chaudhary

Johari

2020

Int J Communication

View full text Add to dashboard Cite

Summary Routing is a process of selecting a path in a network for delivering a packet from source node to destination node. Successful delivery of a message is a challenge, and therefore, this paper proposes an algorithm for a wireless network called Optimized Routing in wireless networks using Machine Learning (ORuML), which uses machine learning algorithm namely, K‐nearest neighbor (KNN), Support Vector Machine (SVM), and Multinomial Logistic Regression (MLR), to predict the network type of the source and destination nodes. The ML model is trained by using characteristic features of a node collected in real time such as battery power utilization, available internal storage, IP address, and range of a node. Intuitively, the MLR should outperform KNN and SVM in terms of accuracy and Area under ROC Curve (AUC). The proposed algorithm determines whether the source and destination nodes are co‐located and also, determines the best neighboring hop for efficient routing.

show abstract

“…These classification models can be formed using an If-Then rule, decision tree, or neural network (Han, Pei & Kamber, 2006). Some of the classification algorithms are ID3, C4.5, Bayesian Network, K-Nearest Neighbor (KNN), and SVM (Soofi & Awan, 2017).…”

Section: Classification Algorithmsmentioning

confidence: 99%

One-Class Conditional Anomaly Detection Algorithm (OCCADA) for Multiple Linear and Logistic Regression

D¹

2020

IJATCSE

View full text Add to dashboard Cite

Traditional one-class anomaly detection is error-prone, leading to numerous false-positive and false-negative anomalies within a context or condition. These anomalies or outliers tend to decrease the accuracy of the Multiple Linear Regression (MLR) and Logistic Regression (LR) models. Also, many real application scenarios have datasets consisting of normal data points or events without anomalies. This study introduces an algorithm using semi-supervised learning on a one-class dataset for classifying conditional anomalous instances. Secondary ground-truth datasets were collected from the Philippine Atmospheric, Geophysical, and Astronomical Services Region XI in Davao City, Philippines Natural Environment Research Council, and Kaggle repositories. The study used six (6) models for evaluation of the One-Class Conditional Anomaly Detection Algorithm (OCCADA) and its application to MLR & LR. The basic assumptions of the models were considered in making valid inferences. The results showed that OCCADA's classification accuracy on continuous and dichotomous response variables were 90% and 91%, respectively. Also, the application of OCCADA to MLR generated a low MAPE value of 0.09 compared to the classic MLR. Similarly, the application of OCCADA to LR increased in percentage of 18.30% compared to the classic LR. The results show that the use of the new one-class conditional anomaly detection algorithm using semi-supervised learning was effective in producing a highly accurate model for classifying conditional anomalous instances and improved the prediction accuracy of MLR and LR models.

show abstract

Classification Techniques in Machine Learning: Applications and Issues

Cited by 231 publications

References 62 publications

Use of acoustic emission in combination with machine learning: monitoring of gas–liquid mixing in stirred tanks

Use of acoustic emission in combination with machine learning: monitoring of gas–liquid mixing in stirred tanks

ORuML: Optimized Routing in wireless networks using Machine Learning

One-Class Conditional Anomaly Detection Algorithm (OCCADA) for Multiple Linear and Logistic Regression

Contact Info

Product

Resources

About