Knn

Fault detection and prediction in the semiconductor industry will enable organizations to optimize their resources and maintain a high production yield for semiconductors. However, the data generated from the sensors are producing an abundance amount of data being kept in the repository and it is difficult for process engineers to analyze and make use of the dataset to make timely and accurate fault prediction and detection. Hence, this research will focus on developing a predictive model to predict and detect faulty equipment in the semiconductor manufacturing process. The quality of the data is enhanced through various analyses, feature selection and dimensionality reduction techniques. Random Forest, Logistic Regression, Support Vector Machine (SVM), K-Nearest Neighbor (KNN) and Artificial Neural Network (ANN) will be the main algorithms used to develop the predictive models and their respective results compared to determine the best model to be used.

The proposed work is about offline recognition of digits and characters using machine learning. MNIST dataset has CSV file. From this, image is obtained by reading it. Then it is preprocessed using RGB converter into binary form that contains some noise, this noise is removed using median filter and then convert into filtered image. Some further pre-processing is done on the image. If the characteristics for one class are distinct and remain constant with variations in this class, then we can call the features as good and desired. KNN Technique is used for classification and recognition of handwritten digits through training and testing. Then, system is made portable by using Raspberry pi and text is converted into speech signal.

The target of SLR or sign language recognition is to interpret the sign language into text, respectively. So the deaf and mute people can communicate with ordinary people easily. Sign language recognition has a tremendous social impact; however, it is challenging due to the significant variations and complexity in the hand actions. There are many existing methods for recognizing sign language that uses handcraft features for describing the motion of sign language and then, based on the features it makes the classification models. To approach the problem, we have discussed considering the KNN that can conveniently extract the features. The proposed model can be validated on a real data set.

Abstract L'Organizzazione Mondiale della Sanità (WHO) ha riferito che nel corso degli ultimi
10 anni le malattie cardiache sono state la prima causa di morte nel mondo. Questo progetto
realizzato con il software Matlab applica un classificatore kNN (k-Nearest Neighbor) per la
diagnosi delle malattie cardiache analizzando l’Heart desease dataset della Cleveland Clinic
Foundation disponibile sul sito archive.ics.uci.edu. Sono state poi analizzate le performance al
variare di k per il classificatore con riduzione delle feature mediante Feature Sub-Set Selection
e con riduzione mediante Principal Component Analysis. Per entrambi gli approcci,
l’accuratezza del classificatore aumenta all’aumentare di k. Anche sensibilità e specificità
aumentano all’aumentare di k. Il classificatore con FSS raggiunge un’accuratezza dell’82%
mentre il kNN con PCA raggiunge un’accuratezza massima dell’80%. In generale si
raggiungono risultati leggermente migliori con l’applicazione della feature subset selection
(FSS) rispetto alla PCA.

Currently, health related issues are the most crucial and life-threatening issues for humans. These health issues include several chronic diseases such as diabetes, cardiovascular disease, respiratory disease, cancer and many more. At present time, cancer is considered as one of the most deadly disease which is increasing consistently in both men and women. Cancer is a medical condition which is caused due to uncontrolled growth of cells present in the lung. The early prediction of its symptoms can help doctors to diagnose it efficiently. Hence, automated computer aided design systems are proposed to identity the lung cancer. These systems are based on the machine learning process where initially, database is learned through their specific process of learning and then classification process is applied. However, these systems fail to deliver the desired performance for medical applications. To overcome this issue, we focus on the deep learning technique and introduced Long-Short Term Memory (LSTM) based model which flows the information in both direction i.e. forward to backward and backward to forward for better learning. This scheme of LSTM is known as BiLSTM because it process the data in two directions. Prior to this, the data preprocessing technique is also incorporated which performs missing value imputation using KNN imputation. The proposed approach is implanted using MATLAB tool and tested on Lung Cancer UCI dataset. The performance is measured in terms of classification accuracy, precision, recall, specificity, sensitivity and F1-score. Further, we compare the performance with existing techniques. The comparative analysis shows that proposed approach achieves better performance when compared with state-of-art techniques.

Currently, health related issues are the most crucial and life-threatening issues for humans. These health issues include several chronic diseases such as diabetes, cardiovascular disease, respiratory disease, cancer and many more. At present time, cancer is considered as one of the most deadly disease which is increasing consistently in both men and women. Cancer is a medical condition which is caused due to uncontrolled growth of cells present in the lung. The early prediction of its symptoms can help doctors to diagnose it efficiently. Hence, automated computer aided design systems are proposed to identity the lung cancer. These systems are based on the machine learning process where initially, database is learned through their specific process of learning and then classification process is applied. However, these systems fail to deliver the desired performance for medical applications. To overcome this issue, we focus on the deep learning technique and introduced Long-Short Term Memory (LSTM) based model which flows the information in both direction i.e. forward to backward and backward to forward for better learning. This scheme of LSTM is known as BiLSTM because it process the data in two directions. Prior to this, the data preprocessing technique is also incorporated which performs missing value imputation using KNN imputation. The proposed approach is implanted using MATLAB tool and tested on Lung Cancer UCI dataset. The performance is measured in terms of classification accuracy, precision, recall, specificity, sensitivity and F1-score. Further, we compare the performance with existing techniques. The comparative analysis shows that proposed approach achieves better performance when compared with state-of-art techniques

Currently, health related issues are the most crucial and life-threatening issues for humans. These health issues include several chronic diseases such as diabetes, cardiovascular disease, respiratory disease, cancer and many more. At present time, cancer is considered as one of the most deadly disease which is increasing consistently in both men and women. Cancer is a medical condition which is caused due to uncontrolled growth of cells present in the lung. The early prediction of its symptoms can help doctors to diagnose it efficiently. Hence, automated computer aided design systems are proposed to identity the lung cancer. These systems are based on the machine learning process where initially, database is learned through their specific process of learning and then classification process is applied. However, these systems fail to deliver the desired performance for medical applications. To overcome this issue, we focus on the deep learning technique and introduced Long-Short Term Memory (LSTM) based model which flows the information in both direction i.e. forward to backward and backward to forward for better learning. This scheme of LSTM is known as BiLSTM because it process the data in two directions. Prior to this, the data preprocessing technique is also incorporated which performs missing value imputation using KNN imputation. The proposed approach is implanted using MATLAB tool and tested on Lung Cancer UCI dataset. The performance is measured in terms of classification accuracy, precision, recall, specificity, sensitivity and F1-score. Further, we compare the performance with existing techniques. The comparative analysis shows that proposed approach achieves better performance when compared with state-of-art techniques.

In this paper we are investigating the possibility of usage of remote sensing satellite data, more precisely sentinel-3 OLCI and SLSTR data, for assessment of bathing water quality. In this research we used data driven approach and analysis of data in order to pinpoint aspects of remote sensing data that can be useful for bathing water quality assessment. For this purpose we collected satellite images for period from start of June till end of September of 2019 and results of in-situ measurement for the same period. Results of in-situ measurement were correlated with satellite images bands and analyzed. We propose a simple method for rapid assessment of possible deterioration of bathing water quality to be used by public health authorities for better planning of in situ measurements. Results of implementation of predictive models based on k-nearest neighbour (KNN) and decision tree (DT) are described. This is an open access article under the CC BY-SA license.

Because of the increasing studies on the big data, holding text as data source, the importance of feature hashing has a major role in the literature now. A usual way of text mining on big data, mostly requires a layer of feature hashing, which reduces the size of fea-ture vector. For example getting the word count yields hundreds of thousands of features in most of the cases and taking the pos-tagging would reduce this number into features about 50. By the feature hashing the size of feature vector reduces reasonably and the data mining processes like classification, clustering or associa-tion can run faster. And in some cases, executing some algorithms are impossible with current hardware, where parallel or distribut-ed programming takes into account. The feature hashing approaches, usually can be categorized into two groups. The first group deals with natural language pro-cessing (NLP) algorithms and tries to extract a relatively smarter hash results, which represents the input ch...

The quantity of text information published in Arabic language on the net requires the implementation of effective techniques for the extraction and classifying of relevant information contained in large corpus of texts. In this paper we presented an implementation of an enhanced k-NN Arabic text classifier. We apply the traditional k-NN and Naive Bayes from Weka Toolkit for comparison purpose. Our proposed modified k-NN algorithm features an improved decision rule to skip the classes that are less similar and identify the right class from k nearest neighbours which increases the accuracy. The study evaluates the improved decision rule technique using the standard of recall, precision and f-measure as the basis of comparison. We concluded that the effectiveness of the proposed classifier is promising and outperforms the classical k-NN classifier.

Since the mid of 1990s, functional connectivity study using fMRI (fcMRI) has drawn increasing attention of neuroscientists and computer scientists, since it opens a new window to explore functional network of human brain with relatively high resolution. BOLD technique provides almost accurate state of brain. Past researches prove that neuro diseases damage the brain network interaction, protein-protein interaction and gene-gene interaction. A number of neurological research paper also analyse the relationship among damaged part. By computational method especially machine learning technique we can show such classifications. In this paper we used OASIS fMRI dataset affected with Alzheimer's disease and normal patient's dataset. After proper processing the fMRI data we use the processed data to form classifier models using SVM (Support Vector Machine), KNN (K-nearest neighbour) & Naïve Bayes. We also compare the accuracy of our proposed method with existing methods. In future, we will other combinations of methods for better accuracy.

In this paper, methods to automatically detect and categorize the severity of skin burn images using various classification techniques are compared and presented. A database comprising of skin burn images belonging to patients of diverse ethnicity, gender and age are considered. First the images are preprocessed and then classified utilizing the pattern recognition techniques: Template Matching (TM), K nearest neighbor classifier (kNN) and Support Vector Machine (SVM). The classifier is trained for different skin burn grades using pre-labeled images and optimized for the features chosen. This algorithm developed, works as an automatic skin burn wound analyzer and aids in the diagnosis of burn victims.

Because of the increasing studies on the big data, holding text as data source, the importance of feature hashing has a major role in the literature now. A usual way of text mining on big data, mostly requires a layer of feature hashing, which reduces the size of fea- ture vector. For example getting the word count yields hundreds of thousands of features in most of the cases and taking the pos- tagging would reduce this number into features about 50. By the feature hashing the size of feature vector reduces reasonably and the data mining processes like classification, clustering or associa- tion can run faster. And in some cases, executing some algorithms are impossible with current hardware, where parallel or distribut- ed programming takes into account.
The feature hashing approaches, usually can be categorized into two groups. The first group deals with natural language pro- cessing (NLP) algorithms and tries to extract a relatively smarter hash results, which represents the input characteristics at maxi- mum or the mathematical hashing algorithms, which do not deal with the context or meaning of the text input and just processes the input for some binary output. For example POS-Tagging ap- proaches can carry on some features of the input to the output on the other hand hashing algorithms like MD5 or SHA-1 has no effect of input, where they only worry about the less collision on the output.
This study focus on the second group of hashing algorithms and criticize the hashing algorithms using Feistel Network which are widely utilized in the text mining studies. We propose a new ap- proach which is mainly built on the substitution boxes (s-boxes), which is in the core of all Feistel Networks and processes the text faster than the other implementations.

Reliable person recognition is integral to the proper functioning of our society.Many researches in face recognition have been dealing with the challenge of the great variability in head pose, lighting intensity and direction, facial expression, and aging. The last few years more and more 2D face recognition algorithms are improved and tested on less than perfect images. However, 3D models hold more information of the face, like surface information, that can be used for face recognition or subject discrimination. A 3D face image is represented by 3D meshes or range images which contain depth information. Range images have several advantages over 2D intensity images and 3D meshes. Range images are robust to the change of color and illumination, which are the causes for limited success in face recognition using 2D intensity images. In the literature, there are several methods for face recognition using range images, which are focused on the data acquisition and preprocessing stage onl...

Precipitation estimation is an important and challenging task in hydrology because of high variability and changing climate. This research involves (1) analyzing changes (trend and step) in seasonal precipitation and (2) estimating seasonal precipitation by disaggregating water year precipitation using a k-nearest neighbor (KNN) nonparametric technique for 29 climate divisions encompassing the Colorado River Basin. Water year precipitation data from 1900 to 2008 are subdivided into four seasons (ie, autumn, winter, spring, and ...

Spam is one of the issues of the Internet which required attention the most especially on the social networking websites and that’s why spam detection is a very concerned issue. Although there are various issues related to the Internet but spam is the one which is faced by everyone who is using the Internet. There are various strategies which are used by the researchers to detect the spam; this is discussed in the Literature Review section of this paper. KNN is one of those techniques and it is abbreviated as K-Nearest Neighbor Classification, it is basically used to find the nearest neighbor based on the pattern recognition and it is a non-parametric technique. Here, in this research paper KNN is used with the Clustering process because of which the process is time efficient and that too with great accuracy. The data which has been provided for the entire process to be performed is extracted from the social networking website Twitter with the help of R package as it provides interface with the Twitter web API (Application Programmable Interface). Various Calculations has been performed to calculate the accuracy, precision, and other parameters and based on these results and respective graphs have been obtained.