The multilayered artificial neural network with learning by back-propagation algorithm configuration is the most common in use, due to of its ease training. It estimated that over 80% all projects development use back-propagation. In algorithm, there are two phases cycle, one propagate input patterns through and other adapt output changing weights network. back-propagation-feed forward can be used many applications such as character recognition, weather financial prediction, face detection...

This paper proposes a speech emotion recognition method based on features and transcriptions (text). Speech such as Spectrogram Mel-frequency Cepstral Coefficients (MFCC) help retain emotion-related low-level characteristics in whereas text helps capture semantic meaning, both of which different aspects detection. We experimented with several Deep Neural Network (DNN) architectures, take combinations inputs. The proposed network architectures achieve higher accuracies when compared to...

Abstract Worldwide, pneumonia is the leading cause of infant mortality. Experienced radiologists use chest X-rays to diagnose and other respiratory diseases. The diagnostic procedure's complexity causes disagree with decision. Early diagnosis only feasible strategy for mitigating disease's impact on patent. Computer-aided diagnostics improve accuracy diagnosis. Recent studies established that Quaternion neural networks classify predict better than real-valued networks, especially when...

Modeling is an abstract representation of real world process. Predicting the likely behavior from observed would be entirely legitimate if relationship were found in data. Two common data mining techniques for finding hidden patterns are clustering and classification analyses. Classification supposed to supervised learning unsupervised with no predefined classes. Clustering tries group a set objects find whether there some between those objects. In this paper we have used numerical results...

Machine Translation (MT) depends upon the MT evaluation for comparing several systems and gives a measure of their efficiency in terms translation sentences. Consequently, has become an integral part procedure it led to development metrics which can automatically assess systems. Lexical like BLEU have been mostly used evaluation. However, these poorly represent lexical relationships impose strict identity matching, leading less correlation with human morphologically rich languages.To...

Compressive strength of the concrete is important for analyzing characteristics concrete. The compressive necessary to know if given mixture meets specified requirements. For sustainability construction, must meet required standards. Machine learning models have been really a handy tool analysis wide range problems. can find pattern or trends in data. purpose paper two folds. First, evaluation performance different machine (regression models) done. In second fold, factors affecting are...

This research proposes a Deep Neural Network architecture for Speech Emotion Recognition called Emoception, which takes inspiration from Inception modules. The network speech features like Mel-Frequency Spectral Coefficients (MFSC) or Cepstral (MFCC) as input and recognizes the relevant emotion in speech. We use USC-IEMOCAP dataset training but limited amount of data large depth makes prone to overfitting, reducing validation accuracy. Emoception overcomes this problem by extending width...

This paper presents a novel approach to exploit the distinctive invariant features in convolutional neural network. The proposed CNN model uses Scale Invariant Feature Transform (SIFT) descriptor instead of max-pooling layer. Max-pooling layer discards pose, i.e., translational and rotational relationship between low-level features, hence unable capture spatial hierarchies low high level features. SIFT captures orientation extracted by Descriptor CNN, therefore, combines feature extraction...

In this paper, a hybrid maximum depth - k Nearest Neighbour (hybrid MD-kNN) method for real time sensor node tracking and localization is proposed. The combines two individual location hypothesis functions obtained from generalized kNN methods. are themselves multiple sensors measuring visible light, humidity, temperature, acoustics, link quality. hybridMD-kNN therefore the lower computational power of outlier rejection ability to realize robust method. Additionally, does not require...

Node.js is a fast green and scalable web development tool that simplifies by allowing developers to use the same language libraries across entire utility stack its non blocking event driven architecture allows for easy creation of packages can handle wide variety concurrent users or connections node js bundle supervisor npm encourages modular improving maintainability speed it ideal micro services architectures decomposing complex programs into smaller feasible offerings be independently...

Failure studies of gear shifter tie rod assemblies are crucial for enhancing automotive system reliability and safety. The origin a assembly failure is analysed through chemical, mechanical, microstructural, fractographic analysis in this paper. Optical imaging reveals the microstructure heat-treated steel, while fractography shows fatigue striations, indicating progressive fracture. According to analysis, penultimate thread likely had fracture due minimal radial space between external head....

In the recent years online social media platforms has been flooded with hateful remarks such as racism, sexism, homophobia etc. As a result, there have many measures taken by various to mitigate spread of hate-speech over internet. One particular concept within domain hate speech is instigating hate, which involves provoking hatred against community, race, colour, gender, religion or ethnicity. this work, we introduce \textit{ProvocationProbe} - dataset designed explore what distinguishes...

We propose a comprehensive end-to-end pipeline for Twitter hashtags recommendation system including data collection, supervised training setting and zero shot setting. In the setting, we have proposed compared performance of various deep learning architectures, namely Convolutional Neural Network (CNN), Recurrent (RNN) Transformer Network. However, it is not feasible to collect all possible hashtag labels train classifier model on them. To overcome this limitation, Zero Shot Learning (ZSL)...

We propose a Multi-Task Learning (MTL) paradigm based deep neural network architecture, called MTCNet (Multi-Task Crowd Network) for crowd density and count estimation. estimation is challenging due to the non-uniform scale variations arbitrary perspective of an individual image. The proposed model has two related tasks, with Density Estimation as main task Crowd-Count Group Classification auxiliary task. helps in capturing relevant scale-related information improve performance comprises...

Underground water and supply of many local places are dirty polluted, due to which good quality is not supplied houses.Humans use techniques enhance the quality, but most need electricity run.Solar distillation one purification that by using direct solar energy.This paper covers experimental analysis enhancement Photovoltaic-Integrated Triple Slope Solar Still (PV-TSSS).Total dissolved solids (TDS), pH value, Electrical conductivity (EC) were measured as 68 -71 ppm, 7.3 -7.4 pH, 44 -45 µS/...