Search

Agriculture is considered to be the driving force of the Indian economy. Production of crops is considered to be one of the complex phenomena as they are influenced by the agro-climatic parameters. From novice to experienced farmers, at times, fail to figure out the suitable crop for their lands, leading to financial loss. This is because of the dynamic change in soil nutrient levels and climatic conditions. Hence, it is important to predict crops according to the presence of the nutrients in a land. Recommending the crops to a farm after considering the nutrients levels of the soil and predicting the yield will largely help the landowner in taking necessary steps for marketing and storage in the future. These results will further assist the industries to plan the logistics of their business who are working in partnership with these landowners. In this work, pH and other soil nutrients are estimated from the input ortho images to recommend crops that can grow well under the given circumstances.

The growth of social networking sites has been tremendous in the last few decades. It has emerged as a platform to share one’s thoughts and opinions and also interact with new people every day. It is possible to predict a person’s emotions by analysing their tweets and posts over a period of time. The main objective is to detect depression among twitter users by performing a sentiment analysis of their tweets. Real time tweets are extracted from Twitter and detected for negative emotions using a lexicon-based ensemble method and a novel Neutral Negative Scoring algorithm. The user history of potentially “depressed” people is obtained and examined by training a neural network to confirm the onset of depression. The depressed people will be given recommendations for initiating positive actions using a recommendation system. It is found that a Bidirectional Long-Short Term Memory network has the highest accuracy of 90% in detecting users with depression.

Neural Networks is applied for solving diverse number of problems involving classification, clustering and prediction of outcomes that forms the basis for resolving Artificial Intelligence related problems. Particle Swarm Optimization is a widely used algorithm to enhance the effectiveness of neural networks. As a part of this article, a new proposition of varying the inertia weight component during run – time of PSO Iterations is being introduced so as to avoid premature convergence in hybrid - PSO algorithm. The experimental results are compared against the feedforward network with random weights and feedforward network with optimised weights using PSO. The results show a significant increase in the performance.

Background: Computer-aided diagnosis (CAD) systems for the detection of lung disorders play an important role in clinical decision making. CAD systems provide a second opinion to the physician in interpreting computed tomography (CT) images. In this work, a CAD system to diagnose pulmonary hamartoma nodules from chest CT images is proposed. Methods: Segmentation of lung parenchyma from CT images is carried out using Otsu’s thresholding method. Nodules are considered to be the region of interests (ROIs) in this work. Texture, shape and run length based features are extracted from the ROIs. Cosine similarity measure (CSM) and rough dependency measure (RDM) are used independently as filter evaluation functions with ant colony optimization (ACO) to select two subsets of features. The selected subsets are used to train two classifiers namely support vector machine (SVM) and Naive Bayes (NB) classifiers using 10-fold cross validation. All the four trained classifiers are tested and the performance measures are estimated. Results: CT slices of patients affected with pulmonary cancer and hamartoma are used for experimentation. From the lung parenchymal tissues of 300 CT slices, 390 nodules are extracted. The feature selection algorithms, ACO-CSM and ACO-RDM are run for different feature subset sizes. The selected features are used to train SVM and NB classifiers. From the results obtained, it is inferred that SVM classifier with the feature subsets chosen by ACO-RDM feature selection approach yielded a maximum classification accuracy of 94.36% with 38 features. Conclusion: From the results, it can be clearly inferred that selecting relevant features to train the classifier has a definite impact on the performance of the classifier. Keywords: Computer aided diagnosis, Pulmonary hamartoma, Ant colony optimization, Cosine similarity, Rough dependency, Support vector machine

Discretization is one of the commonly used data preprocessing technique to improve the efficiency of the knowledge extraction process on clinical data. Generally, clinical data contains numeric attributes with continuous values. Data discretization simplifies the original data by transforming continuous data attribute values into a finite set of intervals. Although discretization is capable of handling continuous attributes on clinical data, there are cases where discretization is not an appropriate technique for handling continuous attributes. There are instances where attribute values are vague, imprecise and have multiple distributions with different classes, which challenges the process of mining in clinical data. Hence, there is a need for fuzzy discretization to pre-process the clinical data before mining. The aim of this study is to derive fuzzy discretization from crisp-interval discretization using geometric approach for constructing fuzzy sets, where overlapping region between the fuzzy sets is represented as geometric area. This study comprises of three steps: First, non-overlapping fuzzy sets are constructed using intervals generated from crisp-interval discretization. Second, area of overlapping between the fuzzy sets is computed based on the geometric approach and an average area of overlapping is estimated. Third, fuzzy sets are redesigned based on the estimated average area of overlapping. Fuzzy discretizations for three, five and seven intervals have been examined using Pima Indian Diabetes dataset (PID) and Bupa Liver Disorder dataset (BLD) taken from the University of California Irvine machine learning repository. The variation in performance of crisp and fuzzy discretization methods is measured using six classification approaches namely, tree based approach, probabilistic induction based approach, rule-based approach, network learning approach, kernel-based approach and distance-based approach and a rule-based fuzzy inference system. The results show that the classification accuracy remains stable with less deviation across different classifiers with varying intervals.

The increase in number of accidents is a dangerous situation that needs to be mitigated. If the data from recent past is taken into account, one can see that the number of road accidents has grown exponentially. In fact, roadways record the highest number of accidents in comparison to other modes of transport. Although drowsy driving is not the only contributor, it remains the principal or major issue and can pose a grave threat if it is not averted. This paper thus aims to introduce a method to reduce the number of accidents due to drowsy driving. Using an Electroencephalograph (EEG) headset which consists of multiple EEG sensors and inbuilt Bluetooth transmitter, brain wave data is transmitted to the system. The collected data is fed as input to the prediction model which decides whether the driver is being drowsy or not. The prediction model is trained by using the Auto Regressive and Integrated Moving Average (ARIMA) time series algorithm to predict whether the person is being drowsy. If so, the proposed system can be used to trigger an alarm/warning mechanism.

A healthy lifestyle in people is achieved by having balanced and nutritional food. In today's world we do not know with absolute certainty what foods can be consumed and what cannot be consumed, that is, we do not know for sure what foods have good nutritional value and what foods do not. The nutritional facts label is printed on food products all over the world and they are represented using a similar structure. These nutritional facts give data about some of the major nutrients present in the food product such as carbohydrates, protein and so on. These nutrition fact labels are not easily understood by common people. People who are careful about their diet such as those who exercise and diet regularly, trainers, and nutritionists may understand these nutritional facts, but not the common people. To make this information accessible in an easier way by classifying these food products into five levels of healthiness ranging from very healthy to very dangerous is the aim of this project work. This is done by a sequential process of data retrieval, data cleaning, data labeling and supervised learning.

The rating systems are important to find the quality of the product or service. These rating systems serve as a guide in finding the perfect one based on different user criteria. Therefore, a sentiment analysis system has been designed for automatic restaurant rating which will be useful to the people in picking their favorite restaurant. The sentiment analysis for restaurant rating system rates the restaurant depending upon the reviews given by the users. The system breaks user comments to check for sentiment keywords. Once the keywords are found, it associates the comment with a sentiment rank. Sentiment analysis can also be extended further to improve the business process. With the customers' reviews, one can understand the changes in the market and improve their product/service. It is also scalable to any type of environment.

In this work, a computer aided diagnosis (CAD) system to improve the diagnostic accuracy and consistency in image interpretation of pulmonary tuberculosis is proposed. The lung fields are segmented using region growing and edge reconstruction algorithms. Texture features are extracted from the diseased regions manifested as consolidations, cavitations and nodular opacities. A wrapper approach that combines cuckoo search optimisation and one-against-all SVM classifier is used to select optimal feature subset. Cuckoo search algorithm is implemented first using entropy and second without using entropy measure. Training is done with the selected features using one-against-all (SVM) classifier. Among the 98 features extracted from the diseased regions, 47 features are selected with entropy measure giving 92.77% accuracy. When entropy measure is not used, 51 features are selected giving 91.89% accuracy. From the results, it is inferred that selecting appropriate features for training the classifier has an impact on the classifier performance.

Approach to diagnose pulmonary bronchitis from its explicit manifestations on lung CT images.CAD system assists physicians in diagnosis and treatment as a second opinion.Ant colony optimization with tandem-run strategy to select relevant features for classification.Classification of bronchitis using SVM classifier. Background and objectivesComputer-aided diagnosis (CAD) plays a vital role in the routine clinical activity for the detection of lung disorders using computed tomography (CT) images. It serves as a source of second opinion that radiologists may consider in order to interpret CT images. In this work, the purpose of CAD is to improve the diagnostic accuracy of pulmonary bronchitis from CT images of the lung. MethodsLeft and right lung fields are segmented using optimal thresholding from the lung CT images. Texture and shape features are extracted from the pathology bearing regions. A hybrid feature selection approach based on ant colony optimization (ACO) combining cosine similarity and support vector machine (SVM) classifier is used to select relevant features. Additionally, tandem run recruitment strategy is included in the selection activity to choose the promising features. The SVM classifier is trained using the selected features and the performance of the trained classifier is evaluated using trivial performance evaluation measures. ResultsThe training and testing datasets used in building the classifier model are disjoint and contains 200 CT slices affected with bronchitis, 50 normal slices and 300 slices with cancer. Out of 100 features extracted from each CT slice, a subset of 60 features is used for classification. ACO with tandem run strategy yielded 81.66% of accuracy whereas ACO without tandem run yielded an accuracy of 77.52%. When all the features are used for classifier training without feature selection algorithm, an accuracy of 75.14% is achieved. ConclusionFrom the results, it is inferred that identifying relevant features to train the classifier has a definite impact on the classifier performance.

Manual attendance has been carried out for many years which is time consuming and also provides erroneous results. The faculty has to maintain the attendance record in registers and file using pen and paper. With the development in technology, we have to look for an automated way for managing attendance. Using mobile phones provide an alternative way in this direction. The proposed system helps the teaching faculty members to maintain attendance of each class they handle. The attendance is recorded using voice automation and then updated in the database maintained for the attendance. The system also enables the teacher to mail periodic reports of the student to their parents. Thus it bridges the communication gap between the faculty members and the parents.