2821 research outputs found
Sort by
Accuracy prediction of rainfall using Decision Tree algorithm and Random Forest
No full textClimate change has made accurate rainfall forecasting more difficult than ever. In this paper,
an automated tool is developed to obtain and generate information from an online resource. The
decision tree algorithm is used to predict rainfall based on historical climate data. The classification and
regression tree (CART) approach is employed to this result, producing a better accuracy rate. The
algorithm is capable of determining the probabilities of rain on any given day, making it an ideal choice
for various applications involving large dataset
Colour trend analysis using machine learning and histogram
No full textA significant aspect of colour prediction is the process of detecting a
colour palette that represents a collection at a fashion show. People do this
manually, but often with too many images, it becomes a hard task. An automated
machine-learning method has been developed to generate colour palettes for
fashion shows. the model was obtained by fashion pictures dataset, representing
each of 48 images of her from a particular fashion show. This work can be
extended to analyse millions of images from social media feeds and provide data driven insights for colour prediction
Efficient online medical store finding and availability of medicines using Decision Tree compared with Random Forest for improved accuracy
Full text linkMaterials and Methods: Both the Decision Tree (N=10) and
Random Forest algorithms (N=10) were iterated 20
times with different test sizes for the Online Medical
Store Finding And Availability Of Medicines and their
accuracies where noted. The dataset used for this
experimental research consists of 501 records.
Results: Decision Tree is substantially more accurate
(91.47%) than Random Forest (86.45%). The
statistical significance of the Online Medical Store
Finding And Availability is (p<.005 Independent
sample T-test) and This score indicates that the study's
results are statistically significant.. Conclusion:
Compared to Random Forest, the accuracy
performance parameter of the Decision Tree looks to
be greater
Design and implementation of a secure patient recommender and prediction system
No full textEarly disease prediction can help sick persons determine the severity of the disease and take quick action, thus,
a healthcare recommended system is viewed as additional tools
to help patients control and manage their ill-health. Medical
recommended system which provides users with quick and
optimal disease predictions has been in existence for a while;
however, it is faced with several data security issues. Sometimes,
patients confidential data which are stored of the archive after
each recommendation may be accessed by unauthorized persons,
and this can warrant a serve data breach and disclosure of
private medical information. Thus, the focus of our project
is to design a privacy-aware recommended system that not
just makes facilitates quick and easy recommendation for sick
persons but also securely protects stored medical information
from unauthorized access. This system will be designed to
support quick search, recommendation septimal confidentiality
and integrit
Customized CNN with Adam and Nadam optimizers for emotion recognition using facial expressions
Full text linkPeople communicate using one of the
communication types of facial expressions to express their
emotions. Human feelings are detected through facial
expressions to interpret their present state of mood. It stimulates
researchers to work in the field of emotion recognition. The
design of deep learning models is essential to interpret the
human current mind state by capturing the pattern of the facial
gesture through their facial expressions. This study proposed a
customized Convolutional Neural Network (CNN) with various
optimizers Adaptive Moment Estimation (Adam) and Nesterov -accelerated Adaptive Moment Estimation (Nadam) to improve
emotion recognition using the dataset FER-2013. The
customized proposed model is designed by varying the number
of convolution layers, filters, filter sizes, and optimizers. The emotions are recognized using softmax activation in the output
layer. The experimental results have proved that the proposed
model classified the facial expressions with accuracy of 0.841,
0.826 using Nadam and Adam optimizers respectively
Remote monitoring system using slow-fast deep convolution neural network model for identifying anti-social activities in surveillance applications
Full text linkRemote monitoring is the process that monitors and observes information from a distance utilizing sensors or electronic types of equipment. Remote monitoring is used in real-time applications like traffic, forest, military, shops, and hospitals to determine abnormal activities. Earlier research has done video processing methods based on computer vision techniques, but the computational complexity regarding time and memory is high. This paper designs and implements a novel Slow-Fast Convolution Neural Network (SF–CNN) to identify, detect, and classify abnormal behaviours from a surveillance video. The proposed CNN architecture learns the video frames automatically, obtains the most appropriate properties about various objects' behaviour from a large set of videos. The learning process of SF-CNN is carried out in two ways, such as slow learning and fast learning. The slow learning process is enabled when the frame rate is less, and the rapid learning process is enabled when the frame rate is high. Both the learning processes learn spatial and temporal information from the input video. Different objects, such as humans, vehicles, and animals, are detected and recognized according to their actions. All the videos have normal and abnormal activities that vary in various contexts. The proposed SF-CNN architecture provides an end-to-end solution to dealing with multiple constraints abnormal movements. The experiment is carried out on several benchmark datasets, and the performance of the SF-CNN architecture is evaluated. The proposed approach obtained 99.6% of accuracy, which is higher than the other existing techniques
CIA in data security
Full text linkCloud computing is a powerful tool that has enabled access to information and
opportunities on a global level. With the advancements in technology and the
development of smart phones for communication, it is essential for organizations to invest in robust cyber security solutions for their mobile devices. The
CIA triad has been used to protect data in the cloud. This research is aimed at
analyzing the application of CIA principles in protecting data. The framework
for the cloud storage system proposed in this research uses a combination of
asymmetric and symmetric keys together with RSA and AES encryption algorithms to transfer data among users in a safe cloud system. When compared to
the existing model, the accuracy was increased by more than 8% (AES). This
research indicates a system to attain high accuracy in cloud data protection
Compare and contrast two different methods of data collection and analysis from a critical realist ontology perspective in studying my lived experience of conversion therapy and disfellowshipment.
Full text linkWhen managing a research project, it is imperative to decide which ontological view the
researcher-participant stands. Is the world viewed in absolute terms thereby having a positivist
view only believing the observable constructions of the mind. From this stance then helps to
decide epistemology informing both data methods of collection and analysis. This informed
the lived experience researcher to follow logical and correct processes to uncover the post
traumatic grief experience after gay conversion therapy. To best understand the lived
experience self-studies are growing in popularity with the experience standing up to the rigor
of peer review. A dual approach is taken to use autoethnography to explore the church culture
via the lens of fundamentalism, working class, gay lens, and auto-hermeneutics to address the
conversion and disfellowship phenomenon. With the focus of trying to get ideas, attitudes, felt
senses of grief and trauma following conversion therapy to curb suicide, depression, and
minority stress
A positive approach to recovery from drug and alcohol addiction
Full text linkBackground: Addiction is a major public health concern, with risk of significant relapse.
Traditional treatment modes look at correcting deficit, not developing positive personal utility.
Understanding addiction recovery as a process of change and growth, as well as how positive
interventions can improve recovery outcomes are vital to addressing this health concern.
Aims: Contribute valuable and original knowledge to what constitutes successful addiction
recovery. Specifically, in how it can be understood through a positive lens, how this knowledge
can be used to support its efficacy, and how positive interventions can be used to safeguard its
future. The overarching aim is to empirically improve addiction recovery, through a series of
complementary and reinforcing studies that endorse and facilitate it as a state of improved
wellbeing, where recovery is strengthened, and a foundation for future flourishing established.
Method: The conceptualisation of a new recovery model, G-CHIME, which considers growth,
connectedness, hope, identity, meaning in life and empowerment as central to addiction
recovery. This is used with apposite contributions from positive psychology (PP) to study
addiction recovery using a mixed-methods approach. This benefits from the methodological
pluralism advocated in third wave PP, and the philosophical backing of second wave PP, which
recognises growth as a facilitator for positive outcome from a negative life event. To this
effect, a qualitative study comprising (n=15) individual narrative analysis studies aggregated
using G-CHIME as a connecting theory, explores the phenomenon of addiction recovery
through accounts of lived experience. A case study investigates how the Values in Action
(VIA) character strengths model, an important contribution from PP to positive functioning,
can be used to identify and explain (subjectively and objectively) positive traits and capacities
that support recovery. A complementary group study (n=100) analyses the VIA character
strengths of people in addiction recovery, which are most and least represented, how this differs
from normative data and why this may be important. From this, G-CHIME and the VIA
character strengths model are used as inputs to the design of a new treatment programme called
Positive Addiction Recovery Therapy (PART). To study the efficacy of PART, two studies
using a within-subjects design assess its effect on wellbeing, recovery capital, and flourishing.
Inherently this includes the quantitative study of G-CHIME, complementing the qualitative
narrative analysis study using this model. The first is a pilot study (n=30) field testing the
programme, and the second, a replication and follow-up study (n=35), substantiating its
findings. PART is then engineered for eHealth using a novel implementation framework, so it
may reach a wider audience via a website. A user evaluation study (n=20) assesses its
perceived impact and reported quality in comparison to other eHealth solutions using
independent summary data, to gauge its success. To complete the digital branch of this work,
the use of chatbots in addiction is systematically reviewed, the output of which is employed in
a user-led design showcasing a novel addiction recovery chatbot, which is not subject to the
prevalent concerns raised in the review. Criterion-based purposeful sampling was used to
recruit participants in addiction recovery for all qualitative and quantitative investigations in
this research. Results: The narrative analysis studies showed the G-CHIME model is helpful
for understanding addiction recovery, and that the important elements of growth,
connectedness, hope, identity, meaning in life and empowerment were identifiable in each of
the accounts. The VIA character strengths model was found to be effective in identifying
personal assets that benefit recovery, and that these could be subjectively explored and
objectively measured. The character strengths profile of people in addiction recovery was seen
to have two characteristics unique to this population, humour, and teamwork. Both the PART
pilot, and replication and follow-up studies yielded statistically significant results affirming the
positive effect that PART has on wellbeing, recovery capital and flourishing. The PART
website was well received by representative users, who reported higher scores than seen in other studies evaluating eHealth implementation. A perceived impact on health-related change
was also reported. Chatbots were found to be a poorly used resources with serious ethical
concerns, requiring better design. Discussion: The findings from a methodologically diverse
set of studies, including both quantitative and qualitative investigations, support a positive
approach to addiction recovery. G-CHIME has been shown as an effective model for
understanding addiction recovery and the components important to its success, providing
evidence that addiction recovery is the positive outcome of a negative experience. G-CHIME
has been found helpful in intervention design, where it provided theoretical input to a
comprehensive treatment programme that was seen to improve the wellbeing and recovery of
the participants who engaged with it, as well as establishing a foundation for them to flourish.
The pluralistic approach advocated in third wave PP, led to the inclusion of the VIA character
strengths, providing further evidence and direction on how PP can be used with effect to aid
positive function in addiction recovery. The eHealth interventions highlighted the need for
new and efficacious design approaches, cognisant of the target population, which can
disseminate positive interventions to a wider group than achievable through face-to-face
intervention. A methodological limitation in this work means a control group has not been
considered. Future study, using a control group design would advance the credibility of the
conclusions drawn in the quantitative aspects of this work. Implications: This work generated
several contributions to knowledge with implications for practice, policy, and research. PART
has been operationalised in a manual, and funding decisions that support its accessibility for
future service users have been made. It is envisioned this could extend to community referrals
from outside of the service where it is currently delivered. The G-CHIME model and the
analytical approach used in the narrative analysis studies continues to be employed in a curated
series of addiction recovery stories in a peer reviewed journal, further developing its evidence
base as an effective model for studying addiction recovery. The opening studies on the VIA
character strengths model in addiction recovery, sets a foundation for further research on how
it may benefit people in addiction recovery. Similarly, the systematic review offered a starting
study on chatbots in the field of addiction, which could help advance their applied use in
addiction and recovery, so that it does not fall behind other areas of healthcare, as is currently
the case
Lung histopathological detection using image classification
No full textEarly detection of lung histopathology has become crucial and essential for humans. Rapid recognition gives many patients the greatest chance of recovery. Histopathological graphics of biopsy samples tissue from possibly infected areas of the lungs are used by doctors to best solution. The multiple type of lung disease is frequently misdiagnosed and prolonged to detect. The characteristics used to detect Lung Histopathology are extracted from Computed Tomography (CT scan) images. Deep Learning (CL) is a novel method that enables us to improve result’s precision. In this paper, we create DL model to determine the type of lung cancers from Computed Tomography images. Convolutional Neural Networks (CNN) which recognize and categorize lung-cancer type within improved efficiency and less amount of time, that is crucial to deciding on the best treatment approach for patents and their risk of mortality. This paper proposes a tri-category classification which applies to images of lung cancer. Benign, adenocarcinoma and squamous cell carcinoma is performed by utilizing VL, VGG-16, and Le-Net to process an image of lung tissue and obtain functionalities effective for diagnostic techniques. Further, the paper analyzes how handcrafted characteristics can be extracted from raw pictures after various processing steps, and finally, the Python framework (Django) is used to deploy the model. The purpose of using this technique is to obtain some characteristics primarily relevant to lung histopathology forecasting