Browse
Recent Submissions
Item Cancer type driver classification accuracy using spark ML technology(UMT, Lahore, 2021) Hafiz Abdullah TanweerIn this paper, analysis of genes extracted from the body has been performed that can be a driver of tumor, resulting In cancer of different types like breast cancer etc. motivated by the BIGBIOCL The Classifier with Alternative and Multiple Rule Based (CAMUR) Is a core algorithm that Is applied here to dissect large datasets. For the purpose of acquiring the desired goal. Apache Spark as well as MLlIb are used on a stack of Hadoop In local mode. The practice has been performed using the decision tree as well as a random forest. As far as the deployed data Is concerned. In terms of measurement of F and efficiency, random forest has shown better results. For the objective of extraction of genes and other pertinent models, deletion of features has been performed with the deployment of an Iterative algorithm as proposed earlier by CAMUR with a modified version. Finally, the extracted results are facilitated to biologists, so they can analyze whether the extraction Is related or can be a driver of cancer.Item Computationally Technical & Economic Impacts of Electric Vehicles(UMT, Lahore, 2021) Mudasir Ali NawazPeople in Pakistan do not prefer the concept of using electric vehicles over internal combustion engine. In this particular research topic, positive impacts of electric vehicles on Pakistan with special focus on economy and environmental protection have been discussed. Electric vehicles can help to reduce the country's dependence on imported oil. These vehicles can be easily operated by electricity generated from local sources such as coal, hydel and Solar power. Thus, reducing the ongoing impacts of oil import from the debt-ridden economy. On the other hand, Electric vehicles can reduce the noise and air pollution which are the biggest threat to environment. Besides these things Electric vehicles are coming with smart technologies (IOT) such as induction of driver less technology which can reduce the accidents to maximum possible extent. The concept of electric vehicles can also end the monopoly of big fishes who are controlling the oil market by replacing the oil filling stations by state owned and properly organized and regularized private electric charging stations. The other important aspect of electric vehicles is the national security, because almost all the vehicles that are used by armed forces are being run on oil so, by replacing the oil operated internal combustion engines by electric vehicles can serve the bigger purpose of undermining the security. The other important thing is the regularization of vehicles is that currently many electric vehicles are not being exempted from the tax which is not good for any country if you want to have more electric vehicles on roads so by replacing the old vehicles by new smart electric vehicles can help the government. But again, it will not bring any fruits to Pakistan if Pakistan keeps on buying the parts of these vehicles from other countries. The only way to get benefit from this revolutionary concept is to setup the locally established plants for the manufacture of such vehicles and by installing more and more charging stations. Comparisons considering environment, energy, economy, cost, batteries, milage & power between EV and ICE also covered.Item AN ADVANCED METHOD OF SUBSTITUTION-BOXES USING RATIONAL CHAOTIC MAP(UMT, Lahore, 2021) Waqar AkbarMany kinds of chaotic cryptosystem have been proposed. This proposed technique provides an innovative approach for building secure substitute-boxes cryptographically using rational chaotic mapping. The suggested rational chaotic mapping is effectively capable of mapping the plaintext order to a huge 8 × 8 S-box that meets the bijective function necessities. The usage of rational chaotic map retains the consistency of the S-box design process and was found to be consistent in comparison to other current substitution-box methods which are used to generate S-boxes. The suggested S-box example is retrieved which has been critically analyzed following normal quality parameters involving, strict avalanche criterion, bijection, nonlinearity, linear probability, differential uniformity and bit independence criterion. To figure out its cryptographic high point, the performance results are associated with the recent analyzed S-boxes. The basic examinations support that the suggested S-box development system is significantly creative and powerful to produce cryptographically solid substitute-boxesItem A Blockchain-based Framework for Distributed Agile Software Development(UMT, Lahore, 2021) Zareen KalimDistributed Agile Software Development (DASD) has been widely adopted by the offshore software companies which completely rely on trust, transparency, security, traceability, effective communication, and coordination. However, failing to sustain all these characteristics can lead to project overdue or failure, payment clashes between customers and developers, project deal cancellations and customer dissatisfaction. Most of the current available frameworks and tools solve the communication issue but they do not use blockchain technology in order to overcome other major issues which include trust, traceability, security, and transparency. In this paper, we have proposed a blockchain-based framework named as AgilePlus which integrates a private ethereum blockchain to solve all these challenges by executing smart contracts. These smart contracts execute in testing layer for acceptance testing to verify whether all the terms and conditions set by customers have been achieved. Smart contracts also execute in payment layer to verify the payment requirements of developers and automatically distribute payments between them. AgilePlus also assign penalties to the customers for late or non-payments while developers receive a penalty for each overdue task. Moreover, we have solved the blockchain’s scalability issue by using Interplanetary File System (IPFS) as a secondary off-chain storage which results in the fast performance of AgilePlus transactions. Finally, the experimental evaluation proves that AgilePlus successfully overcomes the issues of transparency, communication, coordination, traceability, security and solves the trust issues between customers and developers in DASD.Item ANALYSIS OF HEALTH LEVEL 7 (HL7) OPEN-SOURCE CONNECTORS FOR MEDICAL RECORD EXCHANGE(UMT, Lahore, 2021) MUHAMMAD ASADPresently, Open Source Software (OSS) is being widely utilized in the healthcare networks. With the plurality of different OSS in healthcare network, integration between two or more medical systems is of prime importance. This triggers a need for software to integrate these various systems. Among these integrators, Health level 7 (HL7) is an interoperability communication standard in health care networks. The aim of this study is to examine the effectiveness of some available open source HL7 connectors. After listing some open source connectors, a technical analysis has been done on these connectors, comparing various characteristics and properties of each connector. Open source connectors have also been compared with each other along the basis of user support features available in these connectors. Many of the projects have not been implemented practically on a real time basis. Just about two-third of the available connectors are implemented and used in a few health organizations. Security issues have not been discussed in any of the available connectors. We have found the mirth connector better as compared to others. Nevertheless, it still requires improvements in terms of security and features. It is concluded that there is a considerable development in the OSS for the healthcare integration engine. Nevertheless, still a wide range of OSS applications is in development to facilitate the consolidation among the hospitals.Item Machine learning and Neural Network Techniques for Disaster Flood Risk Management(UMT, Lahore, 2021) Mansoor Ahmed RasheedFloods or natural hazards are viewed as one of the most dreadful climate based cataclysmic events. These natural hazards pose a great threat since they are unexpected. A few subjective techniques exist in the literature for the prediction of the danger level of floods caused by natural events. This paper presents the use of far-off detecting information, for example, upgraded Shuttle Radar Geography Mission (SRTM), Thematic Mapper Plus (TMP), combined with land and field information in a Geographic Information System climate for assessment of flood hazards with Feiran in Egypt. This street is an imperative hallway for the sightseers visiting here for strict purposes (St. Katherine cloister) and is exposed to visit streak floods, causing substantial harm to man-made highlights. Moreover, investigations have been utilized to appraise the blaze flood hazard levels of sub-watersheds. To start with, drainage attributes are caught by a bunch of boundaries pertinent to the blaze flood hazard. Further, examination between the accuracy of the sub-bowls has been acted to comprehend the dynamic ones. Additionally, a point by point geomorphological guide for the most unsafe sub-bowls has been introduced. Likewise, a guide distinguishing delicate areas is developed for the Feiran–Katherine street. At last, the most affected components flood danger have been examined. This SLR is based on papers ranging from 2015-2021 and provides synthesis on different algorithms and procedures based on artificial intelligence in context of how these techniques assist in early forecasting of floodsItem A MODEL FOR SELECTION OF SHORTEST ROUTE TO ELECTRIC VEHICLE CHARGING STATION(UMT, Lahore, 2021) LARAIB IMRANIn past years, travelling was a little bit difficult and time taking process which was then resolved with the help of GPS system due to which you can take shortest route towards your destination. Many others problems of vehicles was also resolved by introducing Electric vehicles. Electric vehicles are operated through battery instead of petrol yet the drawback is that they need charging points and electric charging station to charge the battery. These charging points are cheap to install but they are slow in speed therefore electric charging stations were introduced. Electric charging station occupy large space and need huge amount of money to install. The only way to overcome this problem is to introduce the method that charge maximum cars at electric charging station. The main objective of this paper is explain that how an electric charging station will charge maximum cars in minimum time. For this process dijkstra algorithm will be used to show the shortest path to userItem A SYSTEMATIC LITERATURE REVIEW ON AGILE SOFTWARE MAINTENANCE(UMT, Lahore, 2021) IMRAN ASHRAF CHUHDARYSoftware maintenance is unarguably the lengthiest and most cost phase in the lifetime of software. But in most of the literature, emphasis has been only on exploration of software development in the context of classic or agile methods. Waterfall based methodologies work on the principle of sequential flow of work while agile methods are iterative and incremental in nature. The research has shown that use of agile methods help producing software, lesser in cost and better in quality. Certain statistics are available to support the claim that agile methods have reduced cost of software development from 20 to 30 percent. The reason behind is iterative and incremental nature of agile methods that involve the user stakeholders during the whole SDLC process. But not much attention has been paid in literature to see the impact of agile methods on software maintenance. This study demonstrates several papers that are very useful in depicting the state of software maintenance by using agile methodologies. Almost 30 papers will be studied and investigated which are conducted after the introduction of agile for software maintenance. It can be concluded that by opting agile methodology for software maintenance there is a visible improvement in the context of maintenance cost and quality of the software. This study will also help the professionals to identify the context of using agile methodology and its pros and cons, types of agile frameworks and the integration between them.Item SENTIMENT ANALYSIS OF HYBRID CARS REVIEWS USING MACHINE LEARNING(UMT, Lahore, 2021) HASSAN ALIThe increase in usage of Oil and Gas in has led to environmental problems such as global warming, climate change and shortage of crude oil. Due to these reasons people around the world have started to use Hybrid Technology automobiles in the daily life. With people using hybrid automobiles in daily life, there is need of a technique which can help provide information about cars that are in accordance with the user's wishes, namely the recommendation system. This study proposes a sentiment analysis-based model to mine the consumer’s attitude and emotion which they expressed in the form of reviews on various automobile websites. There are two main objectives of this study: 1) develop an annotated benchmark corpus manually for the English Language reviews for the sentiment analysis, 2) to assess sentiment analysis methods and techniques using the Ngram, and Support Vector Machine (SVM) and Long Short Term Memory(LSTM). Three experts annotated the corpus in two categorize: positive and negative with Cohen's Kappa score of 0.90. Lastly the proposed model was analyzed by comparing the results of both models SVM and Naive bayes model.Item FRAUDULENT RIDE DETECTION USING MACHINE LEARNING AND DATA MINING TECHNIQUES(UMT, Lahore, 2021) MUHAMMAD USAMA RIAZWe see multiple criminal activities in daily life like bank credit fraud detection, fintech, cybercrime, etc. The Ride-Hailing industry is increasing like Uber and Careem with the same ratio of fraudulent activities; for example, fraudulent rides to achieve bonuses are growing. In our circle, many companies are working and providing rides facilities. Several cheaters and fake riders are penetrating day by day. It is challenging for us to check every ride; differentcompanies offer a bonus to drivers/riders to make money. Some drivers make fake/dummy rides to improve metrics like Number of rides, rating, completion rate, login hours, and acceptance rate. Fake rides damage the marketing budget, and customers/riders also disturb. The usage of ride-hailing services like Uber, Careem, bykea, etc., has received significant attention in recent years, increasing the number of fraudsters attempting to exploit these systems. We present a methodology for detecting fraud in ride-booking systems in this research. The fundamental approach to resolve this problem is an example of anomaly identification. Anomaly detection in GPS measures the distance between one point to another point. The longest distance between two points in GPS shows that there is disconnection. In simple words, if the calculated point value is greater than the threshold, it means there is an anomaly otherwise average. This anomaly detection is not a hard and fast rule to define a GPS error or NOT, but it helps us check route anomaly detection. The suggested framework adapts to the fluctuations in data in the ride-booking environment and identifies fraud with high precision. Currently, Uber is using Relational Graph Convolutional Network methodology for fraud detection. Mainly our focus is to detect Fraudulent and fraud-based rides by using ML and Data Mining techniques.Item A NOVEL METHOD TO DESIGN THE SUBSTITUTION BOX BY AN INNOVATIVE CUBIC CHAOTIC MAP(UMT, Lahore, 2021) ABDUL MUQEETCryptosystems are used to prove the security of confidential information and communication networks. For this purpose, various permutation and substitution processes are used. The process of substitution is performed by S-box. There are number of S-boxes that are developed but these are static or week so there is need to develop the strong and dynamic S-boxes. The modern day cipher use key dependent S-box that are dynamic. In this thesis the novel dynamic S-box is proposed. To determine the strength of proposed S-Box some standard tools are defined named as Bijectiveness, Nonlinearity, Strict Avalanche Criterion, Linear Probability, and Differential Uniformity. The results are compared with existing S-boxes.Item AN EMPIRICAL EVALUATION OF CLASSICAL ALGORITHMS FOR SENTIMENT ANALYSIS FOR URDU TEXT(UMT, Lahore, 2021) Abdul MateenSentiment analysis has become one of the growing research areas related to the processing of natural languages and machine learning. There is a great deal of opinion and sentiment on particular topics available online, enabling many parties to discuss these views, such as clients, businesses and even governments. . The first task is to classify the text in terms of whether it expresses opinion or factual information Classification of polarities is the second activity that distinguishes between the polarities that sentences can bear (positive, negative or neutral). In terms of the English language, the study of natural language texts for the recognition of subjectivity and feeling has been well studied. The work carried out in terms of the Urdu language, on the other hand, remains in its infancy; hence, further collaboration between research communities is needed for them to give Urdu a mature sentiment analysis method. In this area, there are recognized challenges; some are inherited from the nature of the Urdu language itself, while others are derived from the scarcity of instruments and equipment from different sources. . In this research, we will construct a sentiment analyzer for Urdu text using a lexicon of Urdu sent units and their contextual information in a given sentence.Item Context Aware Systems in Internet of Things: A Review, Taxonomy, and Open Challenges(UMT, Lahore, 2021) MUHAMMAD MEHDI RAZAIn the recent era of computing, Internet of Things (IoT) has evolved as a very constructive technology. Internet refers to dynamic and ever-evolving environments. It also generates contextual information which varies in terms of content, usability, quality and complexity. Day-by-day, the number of users are rapidly increasing, so that there is tremendous increase in user’s mobility and unreliable sensor availability in IoT. Hence, there is necessity to dynamically adapt their behavior at run time in the context-aware applications. In this paper, we have carried out survey of various approaches related to Context-aware systems and self-learning techniques in IoT. We have also focused on the need of different self-learning techniques to unravel the openness of IoT environment. The evolution of Internet of Things (IoT) has increased the appetite for the energy efficient wireless infrastructures. ). In terms of their traffic needs, many of the IoT gadgets are typically resource-intensive and heterogeneous. Furthermore, for the various ecological environments, these units must be made flexible. In any case, the current traffic planning and delivery cycle estimates are inadequate to meet the complex quality criteria for variable plant data IoT applications. In particular, they can not be implemented in cases where multi-hop correspondence is needed for IoT use. This paper aims to define an appropriate access name with several hops for the Wi-Fi-based IoT devices in IoT Fundamentals. In accordance with their heterogeneous traffic needs, IoT implementations are immediately addressed and are designed in accordance with the unambiguously weighted quality classes. At this stage, the IoT system understanding is provided and an improvement model is implemented, which depends on the criteria for administrative efficiency and background requirements. In addition, an energy-efficient, conscious traffic planning (EE-CATS) calculation is proposed, where a prediction strategy for subsidence specifies the mixture of the model.Item ANALYSIS OF FAST BATTERY CHARGING METHODS FOR ELECTRIC VEHICLES(UMT, Lahore, 2021) AYESHA ASIFTo decrease the consumption of non-renewable energy resources and the environmental concerns, rechargeable batteries have been widely incorporated in transit industry and in hybrid automobiles. The commercialization increased the involvement of electric/hybrid vehicles on a gigantic worldwide scale. A number of state level incentives have been presented to enhance the usage of electric vehicles. As a number electric /hybrid vehicles increase on the roads, there should be an effective method of fast charging for smooth travelling through hybrid vehicles. In this study a comprehensive analysis of fast battery charging methods for electric /hybrid vehicle has been conducted on the basis of current and voltage supply to the energy storage system. An extensive observation of Neural Network techniques, Genetic Algorithms, Queuing model and machine learning algorithms have taken into account to deploy fast battery charging phenomena.Item Towards Requirement Traceability Model for Improving Change Management Process(UMT, Lahore, 2021) MUZAFFAR MEHMOOD KHAN AWANSoftware traceability is a required component of many software development processes. Advocates of requirements traceability cite advantages like easier program comprehension and support for software maintenance (i.e., software change). Requirements traceability is considered crucial in change management for establishing and maintaining consistency between software development artifacts. It is the ability to link requirements back to stakeholders’ rationales and forward to corresponding design artifacts, code, and test cases. Trace links can significantly support change impact analysis, saves effort and profoundly improve software maintenance quality. More complete traceability decreases the expected defect rate in the developed software. The strong impact of traceability completeness on the defect rate suggests that traceability is of great practical value for any kind of software development project. Traceability model is used to make efficient and effective change request impact analysis. There are several types of research regarding software requirements traceability problem. The main problem of these researches is that they do not cover the trace link consistency problem properly and the existing proposed solutions cannot be applied to the software industry with affordable changes. We began by developing a lightweight extraction approach that allows an accurate and quick extraction of essential requirements (abstract interactions) from natural language requirements and the generation of Essential Use Case models from them to check the consistency. We then used automated traceability support to create trace links between requirements. We have evaluated the framework’s efficacy and performance through case study and survey. The results were positive and showed that the proposed approach can be used to manage the consistency and accuracy of trace links.Item DEEP LEARNING IN THE DIAGNOSIS OF LUNG CANCER A REVIEW, TAXONOMY, AND OPEN RESEARCH CHALLANGES(UMT, Lahore, 2021) AZKA SARFRAZLow-dose computed tomography (CT) scans are widely used to diagnose early cancers. Clinical studies show that low CT scans reduce lung cancer by 20% compared to standard radiography. However, conventional low-intensity CT scans are prone to overuse, high cost, and increased radiation exposure. This paper seeks to address these challenges by developing machine learning and in-depth case studies for automated cancer screening and assessing disease progression. The new split-method approach was first developed using two-sided select methods and machine learning methods. This method is designed to include ring nodes mounted on the ring bar but significantly reduce partition errors. Second, we developed a neural network to classify clean nodes according to non-nodes. The simulation model integrates VGG, residual, and multi-network module design to enhance the dynamics of external collection components and various reception constraints. Third, the semantic convolution network (HSCNN) network is defined to form negative nodule rings. The semantic components, predicted to be equal to the deficit per node, facilitate the definition of this type and the improvement of visual acuity. Finally, the Bayesian design as well as the full-time Markov version have been improved. Correctly progression in many cases of breast cancer. The decision-making process selects the exchange of information about the individual cancer, providing the basis for a special research study. Numerous experiments and results have shown the effectiveness of these experimental methods in improving and enhancing the efficiency of low-frequency CT programs.Item PRIVACY PREVENTION OF BIG DATA APPLICATIONS: A SYSTEMATIC LITERATURE REVIEW(UMT, Lahore, 2021) FATIMA RAFIQBig data refers to data collections that are so vast or complicated that standard data processing programs cannot handle them. The quantity of data created by the internet, social networking sites, sensor networks, healthcare apps, and many other organizations is rapidly rising as a result of recent technological advancements. Big data analytics is a term used to describe the process of investigating huge volumes of complicated data in order to uncover hid-den patterns or find hidden relationships. This research focuses on privacy and security concerns in big data. This work also covers the encryption techniques by taking existing methods such as differential privacy, k-anonymity, T-closeness and L-diversity. A number of privacy-preserving techniques have been created to safeguard privacy at various phases of a large data life cycle (for example, data production, storage, and processing). The purpose of this research is to offer a comprehensive analysis of the privacy preservation techniques in big data, as well as to explain the problems for existing systems. IEEExplore, MDPI, Science Direct, SAGE and Springer were searched with the following search terms Data protection prevention, Big Data analysis, cybercrime, safety and cyber security. The advanced repository option was utilized for the search by the use of the following in the search: "Cyber security” OR “Cybercrime”) AND ((“privacy prevention”) OR (“Big data applications”)), in order to adjust the results.Item A FEATURE FUSION BASED HYBRID APPROACH FOR BREAST CANCER CLASSIFICATION(UMT, Lahore, 2021) FATIMA IFTIKHARA detected type of cancer is breast cancer commonly in women. An estimated one in nine women is diagnosed with breast cancer. It is unfortunate that due to a lack of proper facilities, the diagnosis of breast cancer in patients is being delayed, which is leading to an increase in the possible death rate. Many different statistical methods and Machine Learning algorithms are often employed in the study to make breast cancer detection more accurate. Machine learning (ML) has allowed doctors to achieve remarkable results, and healthcare is using ML-based models to detect breast cancer in women. This allows analyzing the healthcare data and uses the traditional computer-aided detection (CAD) to assess breast cancer. Machine learning has become an accepted clinical practice and allows doctors to evaluate the ML model to detect breasts at an early stage. A major aim is to diagnose patients with breast cancer by analyzing the data of patients and classifying them into two categories, having diagnosis results as Benign "B" or Malignant “M” based upon his/her tumor features i.e. its radius, area, smoothness, texture, and perimeter. In this research different machine learning algorithms random forest, logistic regression, KNN, SVM, decision tree, and MLP are used to classify cancer as either its malignant or benign. The Kaggle data set is used for applying these algorithms to get the best accuracy. So, for all the above-mentioned algorithms MLP is one of the more efficient and accurate algorithms to classify the breast tumor. And here also fitted the matthews_corrcoef for MLP is 0.89% and accuracy score for the random forest is 0.94%.Item A SELF-SUPERVISED DEEP LEARNING AND ITS’ APPLICATION A SYSTEMATIC REVIEW(UMT, Lahore, 2021) MOAAZ ZAIGHUM MSITUnsupervised Learning based on unlabeled data from large-scale or any human-an noted labels. But the self supervised learning emulate the way used by humans to classify data. In the thesis, the various schema and evaluation metrics of self-supervised learning techniques are examined, followed by a study of the most widely utilized datasets such as photos, videos, audios, and 3D data and the presently available self-supervised visual feature learning methods. For both image and video feature learning, quantitative performance comparisons of the examined algorithms on benchmark datasets are shown and discussed at the beginning of this portion of the thesis. Finally, this work concludes with a list of prospective future avenues for self-supervised visual feature learning that has been found. Also, a Survey of some of the most useful Self-supervised apps is discussed.Item Internet of Things (IoT) in Livestock Environment: A Comprehensive Survey(UMT, Lahore, 2021) OSAMA OMAR SOHAILThe Internet of Things (IoT) is an emerging paradigm that is creating a benchmark and scaling new heights by transforming real-world things (objects) into smarter devices. IoT has multiple application domains, including healthcare, smart grid, and agriculture. IoT has revolutionized the agriculture industry by providing smart solutions for precision farming, greenhouse, and livestock in terms of food quantity and quality. In context to the present standing of IoT in the Livestock field, a comprehensive survey has been done specifically for the identification of the most prominent applications used in the field of Livestock. We evaluate the contributions made by academicians, technologists, and researchers who revolutionized the livestock industry by implementing IoT technologies. Based on current research, no such comprehensive survey has been conducted in this domain. A rigorous discussion on IoT network infrastructure used in livestock has presented the implication of IoT layered network architecture, network platform, and topology. Moreover, a list of communication protocols and connections of IoT-based livestock systems with relevant technologies are also explored. In addition, IoT-based livestock monitoring, controlling, tracking applications, sensor-based applications, and mobile applications developed for animals monitoring have also been discussed. Furthermore, the security issues in the IoT-livestock scenario are highlighted. In the end, open research challenges in IoT-based livestock systems have been presented.