Semi-supervised text classification using enhanced KNN algorithm

Defective hard candies are usually produced due to inadequate feeding or insufficient cooling during the candy production process. The human-based inspection strategy needs to be brought up to date with the rapid developments in the confectionery industry. In this paper, a detection and classification method for defective hard candies based on convolutional neural networks (CNNs) is proposed. First, the threshold_li method is used to distinguish between hard candy and background. Second, a segmentation algorithm based on concave point detection and ellipse fitting is used to split the adhesive hard candies. Finally, a classification model based on CNNs is constructed for defective hard candies. According to the types of defective hard candies, 2552 hard candies samples were collected; 70% were used for model training, 15% were used for validation, and 15% were used for testing. Defective hard candy classification models based on CNNs (Alexnet, Googlenet, VGG16, Resnet-18, Resnet34, Resnet50, MobileNetV2, and MnasNet0_5) were constructed and tested. The results show that the classification performances of these deep learning models are similar except MnasNet0_5 with the classification accuracy of 84.28%, and the Resnet50-based classification model is the best (98.71%). This research has certain theoretical reference significance for the intelligent classification of granular products.

show abstract

“…CDNN [36] 76.73% Enhanced k-NN [37] 74.90% SVM [6] 90.98% Random forest [38] 90.33% Resnet-18-based model 98.20%…”

Section: Models Accuracymentioning

confidence: 99%

Detection and Classification of Defective Hard Candies Based on Image Processing and Convolutional Neural Networks

Wang

Chen

et al. 2021

Electronics

View full text Add to dashboard Cite

show abstract

“…Now all the data will be in text format. Textual data is considered unstructured and as result it is difficult to analyze (Wajeed & Adilakshmi, 2009). Analyzing textual involves classifying or categorizing.…”

Section: Categorizing Text Data and Audio Recordingmentioning

confidence: 99%

“…Classification can be flat or hierarchical. In flat classification, categories are on the same level in a parallel format, one category does not supersede another (Wajeed & Adilakshmi, 2009). See Figure 1 for illustration.…”

Section: Categorizing Text Data and Audio Recordingmentioning

confidence: 99%

See 1 more Smart Citation

Towards Harnessing Phone Messages and Telephone Conversations for Prediction of Food Crisis

Lukyamuzi

Ngubiri

Okori

2015

International Journal of System Dynamics Applications

View full text Add to dashboard Cite

Food insecurity is a global challenge affecting millions of people especially those from least developed regions. Famine predictions are being carried out to estimate when shortage of food is most likely to happen. The traditional data sets such as house hold information, price trends, crop production trends and biophysical data used for predicting food insecurity are both labor intensive and expensive to acquire. Current trends are towards harnessing big data to study various phenomena such sentiment analysis and stock markets. Big data is said to be easier to obtain than traditional datasets. This study shows that phone messages archives and telephone conversations as big datasets are potential for predicting food crisis. This is timely with the current situation of massive penetration of mobile technology and the necessary data can be gathered to foster studies such as this. Computation techniques such as Naïve Bayes, Artificial Networks and Support Vector Machines are prospective candidates in this strategy. If the strategy is to work in a nation like Uganda, areas of concern have been highlighted. Future work points at exploring this approach experimentally.

show abstract

“…Lv [25] proposed a semi-supervised learning approach using local regularizer and unit circle class label representation. Wajeed and Adilakshmi [26] proposed a semi-supervised method to increase accuracy of k-NN for text classification. Giannakopoulos and Petridis [27] applied a semi-supervised version of Fisher discriminant analysis together with k-nearest neighbor to a diarization system.…”

Section: Introductionmentioning

confidence: 99%

A Semi-Supervised Approach Based on k-Nearest Neighbor

Liu¹,

Zhao²,

Zou³

et al. 2013

JSW

View full text Add to dashboard Cite

k-nearest neighbor (k-NN) needs sufficient labeled instances for training in order to get robust performance; otherwise, performance deterioration occurs if training sets are small. In this paper, a novel algorithm, namely ordinal semi-supervised k-NN, is proposed to handle cases with a few labeled training instances. This algorithm consists of two parts: instance ranking and semi-supervised learning (SSL). Using SSL, the performance of k-NN with small training sets can be improved because SSL enlarges the training set by including unlabeled instances with their predicted labels. Instance ranking is used to pick up the unlabeled instances that are included to the training set. It gives priority to the unlabeled instances that are closer to class boundaries because they are more likely to be correctly predicted (these instances are called high confidence prediction instances). Thus, SSL benefits from high confidence prediction instances if they are added into the training set early. Experimental results demonstrate that the proposed algorithm outperforms the semi-supervised k-NN and the conventional k-NN algorithms if the training set is small.

show abstract

Semi-supervised text classification using enhanced KNN algorithm

Cited by 9 publications

References 3 publications

Detection and Classification of Defective Hard Candies Based on Image Processing and Convolutional Neural Networks

Detection and Classification of Defective Hard Candies Based on Image Processing and Convolutional Neural Networks

Towards Harnessing Phone Messages and Telephone Conversations for Prediction of Food Crisis

A Semi-Supervised Approach Based on k-Nearest Neighbor

Contact Info

Product

Resources

About