Browse Articles
Discover research articles across all indexed journals
PREDICTIVE ANALYSIS OF JOB RECRUITMENT USING MACHINE LEARNING APPROACH
The main objective of Predictive Analysis for Job Recruitment using Machine Learning Approach project is to create a web portal that can predict the probability of candidate getting job in particular company. This project could help students to know whether they will get job or not and even company can use this project to know the chances of candidate getting job. The placement dataset has been considered, which is taken from Kaggle and the proposed system uses Support Vector Machine (SVM) algorithm to achieve the objective. Some of the Classification methods such as logistic regression, decision tree classifier, support vector machines, random forest classifier, KNN and Naïve bayes are implemented to check the accuracy and efficiency and it is noticed that SVM gives 91% of accuracy. Hence SVM algorithm is used to train the prediction model. Input to the system will be a text/query given by the candidate or a student and the resulting output will be a probability of getting job generated by the system.
RAGA DETECTION
The Indian music genre bestows an artist to offer their own particular flavor to a raga makes it delicate for a neophyte to spot two contrasting performances of the identical raga. The importance of Raga identification in Indian music cannot be inflated. The analysis must begin with relating the underpinning raga. There are various attempts made in relating the raga in an exceedingly music. The identification of ragas is intelligible and comes only after respectable quantum of revelation. For automated identification, several the attributes of ragas must be converted into applicable features. This becomes particularly difficult for Indian music because of the subsequent reasons which must be addressed while transforming a music piece into swara strings It might be difficult for a novice to distinguish between two different renditions of the same raga because the Indian music genre allows a performer to add their own unique flavour to a raga. It is impossible to overstate the significance of raga identification in Indian music. Identifying the underlying raga is the first step in the investigation. Various attempts have been made to relate the raga in a very musical way. Raga classification is comprehensible and follows a reasonable amount of revelation. Several ragas' properties must be transformed into relevant traits in order for them to be automatically identified. Due to the ensuing considerations that must be made while converting a musical composition into swara strings, this becomes especially challenging for Indian music. During a performance, numerous instruments are also used to create a musical composition. Indian music uses a relative scale rather than a Kelvin scale for its notes, in contrast to Western music. A very raga lacks a rooted beginning swara. In Indian music, notes contain a spectrum of frequencies (also known as oscillations) surrounding them rather than a fixed frequency. Each raga has an unfixed number of swaras, and if all of the raga's attributes are present, colourful extemporizations are permitted when citing a particular raga. These elements present a significant obstacle to the automatic discovery of raga. One of the methods for raga type is the recapitulation of raga into swaras at each point in time and classification using a classifier similar to K Nearest Neighbor model has been created and published while the model offers more nuance on the test dataset sample. There have been attempts to relate the raga to a piece of music. A strategy for determining the raga type involves resuming the raga into swaras at regular intervals and categorising using techniques like KNN and SVM.
VIDEO SUMMARIZATION
The general public now has access to a vast amount of multimedia information thanks to recent technological advancements and the quick expansion of consumer electronics, making it challenging to effectively consume video material among the thousands of options accessible. By choosing and presenting the most educational or fascinating materials for users, we provide a method to quickly summarize the content of a lengthy video document. The practice of condensing a raw video into a more manageable form without losing much information is known as video summarizing. Either a comprehensive analysis of the full movie or the local differences between neighboring frames are used to achieve this. The majority of such approaches rely on universal characteristics like color, texture, motion data, etc. Video summaries are evaluated depending on the sort of content they are formed from (object, event, perception, or feature-based) and the functionality made available to the user for consumption (interactive or static, personalized or generic). The suggested system analyses each frame of a video as input before producing a summary. Each frame receives a score that is used to compare it to a threshold value in the final phase. Every frame whose frame score exceeds the threshold is chosen as a key frame and is represented in the final movie summary. This technique enables us to condense video information of various lengths while guaranteeing that the key moments are included. The purpose of video summary is to facilitate quick access, speed up browsing through a sizable video database, and offer a condensed video representation while maintaining the core activities of the original video.
SMART DETECTION OF DAMAGED ROADS USING MACHINE LEARNING
Finding potholes is crucial to reducing accidents all throughout the world. Numerous studies have been conducted, however in order to collect sensor data, they need certain instruments or technologies. In this research, we suggest a method for using a smartphone to detect potholes, and categorization is utilizing. Transfer Learning. The experimental finding supports the suggested strategy. Gives us efficiency in terms of performance and implementation. An array of the gathered data was processed using several techniques, and features from various frequencies. In addition to different machine-learning classifiers, domains were retrieved According to the results, characteristics from the temporal and frequency domains performed better than other features at spotting potholes. The Random Forest classifier demonstrated the greatest classification performance for potholes among the studied classifiers, with a precision of 88.5 percent and recall of 75 percent. Finally, we tested the suggested method's resilience and universality using datasets obtained from various types of roads.
SECRET IMAGE SHARING AND IDENTIFYING FAKE IMAGE USING BLOCK CHAIN
The detection of fraudulent images and the attribution of image sources has become a hot topic of debate in both the news industry and social media. Anyone may quickly create or edit digital content in the age of digitalization, and they can then simply broadcast it on social media networks. The use of these social networking sites has created new difficulties for practical application, such as the viral transmission of incorrect or misleading material with harmful intentions. On the one hand, they greatly facilitate modern communication. A currently very emerging technology for data sharing and application is block chain. By using data encryption, timestamps, and distributed consensus, it is possible to exchange decentralised information in distributed systems without relying on one another, which increases the effectiveness of data sharing and application. The massive data remote sensing image system can fully exploit this technology, and the multi-system shared node storage system can be controlled effectively and uniformly to increase the system's economic efficiency. This study suggests important research technologies, builds the shared architecture based on block chain technology, and offers a theoretical framework for non-engineering practise.
VEHICLE VIOLATION DETECTION SYSTEM USING MACHINE LEARNING
There are now several commercially available programmes, such as Google Goggles, that read images and content from the advanced cell that were captured by the PDA's camera. Such applications are subject to the following limitations and challenges, such as: The requirement for an Internet connection; the requirement to extract unique characters on an embedded portable stage, which has requirements as to both memory and processor; and the increasing difficulty of the recognition task due to the presence of specular reflection. The majority of the time that photographs are taken is at night, therefore they may be hidden and have a low level of complexity. Content it is a test to localize from the captured image's background. Therefore, we have tried to detect the number plate using a photo recovery-based methodology. One sophisticated consistent picture recovery framework for mobile phones. In that piece, the authors demonstrated a method for seamlessly fusing the physical and technological worlds using paper records; unfortunately, the same method cannot be applied in our case.
WEED DETECTION AND CLASSIFICATION
In the agricultural sector, weed detection and categorization are of critical technical and financial relevance. It is possible to automate various processes, such as shape, colour, and texture, in weed control systems. Building a machine vision weed control system that can locate weeds in real-time is the aim of this article. The system is designed to categorise photos into wide and narrow classes for on-the-fly spraying of selective herbicides. The developed Edge Link Detector-based algorithm has been put to the test on weeds in various areas, and the results have revealed how good the algorithm is at identifying weeds. Additionally, the outcomes demonstrate a very trustworthy performance on weeds under various field circumstances. Over 240 sample photos (wide, narrow, and no or little weeds) including 100 samples from broad weeds, 100 samples from narrow weeds, and the remaining 40 samples from no or little weeds were classified with over 93 percent accuracy, according to the results analysis.
STUDENT CAREER GUIDANCE
Career Counselling is a method that makes a specialty of assisting one apprehend one’s personal self, in addition to paintings trends, in order that it is easy to take an knowledgeable choice approximately profession and education. After 12th standard, many of the students have an agony about choosing the field they are interested in. Our web-based student career counselling application makes easier for students to choose their career path wisely. The student is required to enter few details of his academics and extra-curricular activities, based on the input of the student; our application provides a proper career path. The Overall idea of this project is to provide a solution for high school pass out students through giving the feedback by considering the student marks and various other factors. Our web-based application not only helps the students but also helps the parents to get the guidance to know the status of their children. In this project we are applying the machine learning algorithm to conduct the task and predict the output based on the given input values. We are using throughout this project, numerous machine learning techniques for predicting the result based on the given input values. After taking the test on our website, the student will have a clear idea for choosing the career wisely through which the failure rate of choosing a wrong career will be decreased.
IMAGE CHARACTER RECOGNITION USING CONVOLUTIONAL NEURAL NETWORKS
Object Recognition is a challenging and exciting task of computer Vision. Object recognition is to describe a collection of related computer vision tasks which involves identifying objects in images. Image classification involves predicting the class of objects in an image. Object localization refers to identifying the location of objects in an image. Image Annotation in Machine Learning is the process of creating bounding boxes around the localized images. Now with the advance of deep learning and neural networks, we can finally tackle these problems without coming up with various heuristics real time. By now, there are many different algorithms and concepts for object detection published. Just a few are lightweight enough to run them on live interface that results in an appropriate frame rate. Object detection networks are for example the You Look Only Once (in short: YOLO) network or the Faster R- CNN network. In this project our aim was to develop a software application that takes an image and produce the Informative links about the objects present in the image. Our aim in this project is to develop a software product that runs the SSD to identify objects in an image and the identified images are searched in Google API to give appropriate informative links of those objects
REVIEW SPAM DETECTION USING MACHINE LEARNING
Reviews, ratings, and experience stories left by customers on e-commerce websites and other online services are helpful to both buyers and sellers. The reviewer can foster more brand loyalty and aid in the understanding of other consumers' product experiences. Similar to how reviews help customers earn more profiles, reviews help businesses sell more things by enhancing customer satisfaction. However, suppliers may, unfortunately, abuse these review processes. For instance, one might fabricate positive evaluations to boost the reputation of a brand or attempt to denigrate rival brands' goods by posting fictitious reviews about them. Utilizing various machine learning techniques and tools are examples of existing solutions with supervision. Unlike previous work, I decided to use a wide range of vocabulary to work with, such as many datasets integrated into one enormous data set, rather than a confined dataset. Based on the reviews' text and emoji usage, sentiment analysis has been implemented. Review fraud is recognised and classed. The use of the Linear SVC, Support Vector Machine, and Random Forest algorithms yields the test results.
RESUME ANALYZER
Resume analysis is the process in which a machine analyses a resume based on given requirements of the job description. With the flood of resumes received by companies, it is not effective and also not possible for a person to go through a number of resumes to select a candidate. They have become very popular among the companies in the process of determining candidate selection. The main objective of the project is to be able to match the requirements and skills from a job description to the resume applied. This gives an instantaneous result on whether the resume is accepted or rejected. The end process allows the company itself to be able to select candidates without the requirement of a third party and thus is also cost effective. A big number of resumes could be used in this project to sort the necessary application using various classifiers. Following classifications, the top n candidates will be sorted in accordance with the job description using content-based recommendation and cosine similarity. The project employs k-NN to determine which CVs are most similar to the supplied job description. Through machine learning, the system evaluates a resume for a particular position using NLP.
TRAFFIC PREDICTION USING MACHINE LEARNING
The goal of this study is to provide a mechanism for forecasting precise and timely traffic flow data. Traffic Environment refers to anything that could have an impact on how much traffic is moving along the road, including traffic signals, accidents, protests, and even road repairs that might result in a backup. A motorist or rider can make an educated choice if they have previous knowledge that is very close to approximation about all the above and many more real-world circumstances that can impact traffic. Additionally, it aids in the development of driverless cars. Traffic data have been growing dramatically in the recent decades, and we are moving toward big data concepts for transportation. The current approaches for predicting traffic flow employ certain traffic prediction models, however they are still insufficient to deal with real-world situations.
PREDICTIVE ANALYSIS OF SPORTS DATA USING MACHINE LEARNING
There are numerous methods for making sports predictions, and data analysis is crucial to predicting. Previous attempts in sports data analysis have resulted in the prediction of sports such as football, tennis next shot location prediction, Olympic athlete performance, basketball slam dunk shot frequency, and many more. Cricket prediction is tough due to the numerous variables that might affect the result or outcome of a cricket match. Previously, simple cricket match prediction systems focused on the venue, ignoring aspects such as weather, stadium size, captaincy, etc. Factors such as the match's location, pitch, weather conditions, first-pitch batting, and fielding all play a role in forecasting the match's outcome. To predict, suitable models are required, and data mining allows the required information to be extracted from data sets. This paper is a review of techniques used for predicting the winners of three different games. In order to anticipate various facts linked to a certain match, such as the outcome of the match, an injured player's performance in the match, the discovery of new talents in the game, etc., various machine learning algorithms can be used to exploit the statistical data of the game. The objective is to correctly forecast the outcome of a specific game.
PREDICTING STOCK MARKET TRENDS USING MACHINE LEARNING AND DEEP LEARNING ALGORITHM
Intention of StockMarket Prediction is to forecast future worth of a company's monetary stocks. utilising machine learning produces forecasts based on the values of current stock market indices Using their prior values as training data is a new development in stock market prediction technology. Predicting the performance of the stock market is one of the most challenging tasks.Prediction involves a huge number of variables, such as the distinction between physical and psychological factors, rational and irrational conduct, and more. Share prices are unpredictable and challenging to forecast accurately as a result of a combination of all these variables. Many researchers have conducted studies on the upcoming market developments movement. Data is a key source of efficiency because stock is made up of dynamic data. The prediction's efficiency has an effect on the same chances. To make accurate predictions, machine learning employs a variety of models. Machine learning techniques have the potential to uncover previously unknown patterns and insights using features like an organization's most recent announcements, quarterly sales statistics, and so forth, which can then be utilised to produce impeccably correct forecasts. The project's goal is to predict stock values using machine learning that is based on long short-term memory (LSTM) and regression. All of the variables—open, close, low, high, and volume are taken into account.
TOMATO LEAF DISEASES PREDICTION
This paper deals with the optimized real time detection of diseases that affect the plant and the area affected using Convolutional Neural Networks (CNN) algorithms. So that appropriate fertilizers can be used to prevent further damage to plants from pathogenic viruses. The activation function is the core of the CNN model as it incorporates the non – linearity to have anauthentic artificial intelligence system for classification. ReLu is one among the best activation functions, but hasa disadvantage that the derivative of the function is zerofor negative values and leads to neuronal necrosis. New mathematical activation function is developed and compared with existing activation functions to improve the accuracy and performance of the system on a Tensor Flow frame work. The experimental results on trained databases show that the developed activation function has improved the CNN model accuracy and performance i.e. 95%. The training speed of the CNN model is improved by 83% when implemented in ARM processor using the proposed optimizer. Further area affected by disease is calculated by using K–means clustering algorithm for optimization of fertilizer usage.
HEALTH CARE CHATBOT
Chatbots mimic human speech to make the software more user-friendly or just for amusement purposes. NLP (Natural language Processing) and AI (Artificial intelligence) are new emerging technologies that can be utilized to improve the capacity of chatbots to stimulate a more natural and free-flowing conversation. Chatbots can be used to offer customers support and services as more and more mobile device users switch to more frequent use of texts and messaging Healthcare chatbots provide a customized experience to each user in ways that may be more practical and effective than humans are capable of. By communicating with consumers in a human like manner, a medical chatbot enhances the work of a healthcare professional and aids in enhancing their performance. Medical chatbots are conversational AI-powered tools that facilitate communication between patients, insurance providers, and health care professionals. These bots can be quite useful in providing timely access to pertinent healthcare information to the appropriate parties. We discuss current chatbot research and uses in the fields of medical and health care education, our main areas if concentration includes the use of virtual patients in medical education patient education regarding healthcare issues, and the use of chatbots as course assistants to improve the curriculum for healthcare professionals. Because of their mobility and affordances, chatbots are increasingly being used into the teaching and learning processes in the field of healthcare education. Clients of this application have discussions with the wellbeing chatbot likewise to how they would have discussions with different clients. Through a series of questions, the health chatbot helps people identify their symptoms in order to assist them decide whether or not to see a doctor. People who are unsure of whether their symptoms are temporary or call for further testing from a doctor may find this.
CYBERBULLYING DETECTION
Cyberbullying is identified as bullying that prevails via digital gadgets such as cell phones, smartphones, laptops, Chrome book, personal computers, notebooks, and tablets etc. It is also known as electronic bullying, is a well-known peril of technological evolution that involve selective and multiple actions at odds with one or more people making use of internet equipment. Unlike "traditional" bullying, cyber bullying does not usually happen only in the park or outside of school. It happens in a variety of cyberspace venues and mediums, but it mostly happens where teenagers mingle. It is more common than most parents and teachers acknowledge. Students frequently attempt to resolve situations on their own origin or either bullies. This could work in the short term. However, the mental health implications of cyberbullying can be detrimental.
SMART IGNITION USING HELMET DETECTION
Over the past few decades, the percentage of individual transportation has increased from 39.2% to 77% globally, which shows rapid growth in the sales of Automotive Industries, as this increases, it also shows an increased rate of road accidents. The goal of this project is to use the Raspberry Pi, for the envisioned advanced system. The Raspberry Pi will serve as a command module. This project aims to introduce an efficient and feasible methodology by implementing Object Detection for improved safety. The project incorporates image processing where the authorized user’s wearing a helmet are recognized and verified with the database and allows users to turn on the Ignition System. Implementation of this system can be done on motorcycles with minimal installation cost which provides safety to the users. Keeping in mind the future perspectives, Smart Ignition Using Helmet Detection can surely provide a better system thereby ensuring passenger safety.
IRIS RECOGNIZATION USING CONVOLUTIONAL NEURAL NETWORK
A Multi Biometric System Based on the Right Iris and the Left Iris Using the Combination of Convolutional Neural Networks, Deep Learning Algorithm. Eyes are the most valuable organ of the human body, which process the images of objects in the vicinity. The eyes are essential and fragile organs at the same time. They are characterized by several complex structures. These structures understood with their atomies and functioning as well as basic physiology. They are subdivided into layers from front to back. The lower and upper eyelids help protect the eyes. The eyes interpret size, shape, colour and distance of the objects and give a 3D picture of the objects visible. The eyes are seated in a hollow cone shaped cavity (socket) of the human skull named orbit. The movements of the eye are regulated by six muscles. Fatty tissues in the orbit surrounding, which protect the eyeball, give it flexibility to function. The ad nexal structures, like eyebrows, eyelashes, eyelids, protect eyes from foreign elements and injuries. The liquids maintaining the shape of the eyeballs are called aqueous humour and vitreous humour. The field of IT security will be completely changed by biometrics. This technology is being used by numerous businesses, governments, and international corporations to secure sensitive data. The biometric passport and driver's permit have both become more prevalent recently. Both academics and Multimodal biometric recognition systems are used by industry professionals to boost security and stability of the system. The usage of dual-mode biometric systems has advantages over single-mode systems, although a single modality alone renders the system weak. Professionals in biometrics choose a multimodal system. Combining the following techniques is possible: voice, iris, fingerprint, and facial recognition, acknowledgment, signature, etc.
INFRARED WIRELESS UNDERWATER COMMUNICATION SYSTEM
The Wireless communication has been booming in the recent years. However, not much has been done to accomplish communication in water in a wireless manner which can be useful in many scenarios. Most of the communication modes that are existing as of now are either not reliable or are very expensive. IR rays can pass through water and thus be used in case of line of sight for communication purpose in this medium. This IR communication using Keyboard uses two units both controlled by ATmega family microcontroller. Both of the units have IR trans-receivers as communicating agents. Keyboards are needed to be connected on both the ends to enter input to the system. The communication messages get displayed on the LCD connected to the system. The system communicates with confirmation key that is sent back by receiving unit to the sending unit. In this way wireless communication is implemented with great efficiency within a line of sight range of about 3-4 meters under the water with the help of IR wireless underwater Communication project.