IRAQI JOURNAL OF COMPUTERS, COMMUNICATIONS, CONTROL AND SYSTEMS ENGINEERING

Cloud computing is an evolving and high-demand research field at the
forefront of technological advancements. It aims to provide software resources and
operates based on service-oriented delivery. Within the infrastructure as a service (IaaS)
framework, the cloud offers end customers access to crucial infrastructure resources,
including CPU, bandwidth, and memory. When a cloud system fails to deliver as
expected, it is referred to as an event, signifying a deviation from the anticipated service.
To meet their service-level agreement (SLA) obligations, cloud service providers (CSPs)
must ensure continuous access to fault-tolerant, on-demand resources for their clients,
particularly during outages. Consequently, finding the most efficient ways to accomplish
tasks while considering the rapid depletion of resources has become an urgent concern.
Researchers are actively working to develop optimal strategies tailored to the cloud
environment. Machine learning plays a critical role in these endeavors, serving as a key
component in various cloud computing platforms. This study presents a comprehensive
literature review of current research papers that employ machine learning algorithms to
propose strategies for optimizing cloud computing environments. Additionally, the survey
provides authors with invaluable resources by extensively exploring a diverse range of
machine learning techniques and their applications in the field of cloud computing. By
examining these areas, researchers aim to enhance their understanding of efficient
resource allocation and scheduling, addressing the challenges posed by resource scarcity
while meeting SLA obligations

The voice signal carries a wide range of data about the speaker, including their
physical characteristics, feelings, and level of health. There are several uses for the estimate
of these physical characteristics from the speech in forensics, security, surveillance,
marketing, and customer service. The primary goal of this research is to identify the auditory
characteristics that aid in estimating a speaker’s age. To this end, an ensemble feature
selection model is proposed that selects the best features from a baseline acoustic feature
vector for age estimation from speech. Using a feature vector that covers various spectral,
temporal, and prosodic aspects of speech, an ensemble-based automatic feature selection is
performed by, first calculating the feature importance or ranks based on individual feature
selection methods, then voting is applied to the resulting feature ranks to attain the top-
ranked subset by all feature selection methods. The proposed method is evaluated on the
TIMIT dataset and achieved a mean absolute error (MAE) of 5.58 years and 5.12 years for
male and female age estimation.

The load forecasting is a human or computational technique for
accurate preanticipation of electrical load to enhance reliable operation and
optimal planning control of system plant for electrical energy flowing without
facing any economical and technical limitations, therefore appropriate
estimation for present and future consumption cost of electrical loads which are
necessary to predict the load demand for generating near to accurate power.
During advanced technology at the last few decades, artificial neural
networks(ANNs) have been extensively employed in electrical system, they are
trained using historical data obtained from plant station. This work is intended
to be a study of short-term load forecasting (STLF) basis for a power predicted
applied to the actual past load data displayed from Azadi station for Feb.2022
were used in training and validation system of neural grid. The result was
evaluated by mean square percentage error of (32.7) for the forecasting
dynamic time series method to solve the data over hours, days, and weeks in
advance, using a kind of non-linear filtering. Short-term load forecasting tried
out with main stages; predicted power load data sets, network training, and
forecasting. Neural network used has 3-layers: an input, a hidden, and an
output layer. The number of hidden layer neurons can be varied for the different
network performance. The active power generation faces economical and
technical challenges, therefore appropriate evaluation of loads are much
needed.

Multi-Object Detection and Tracking (MODT) are essential in many
application fields. Still, many enhancements in the speed of detection and tracking were
required to overcome the challenges during implementation. This paper presents a new
algorithm system for (MODT) to improve the execution time to be robust in real-time
applications. A background subtraction detection algorithm with a Kalman filter was
used to track and predict the object position and speed parameters. To improve the
processing time, its needs to reduce some frames in a way that does not affect the
detection accuracy too much and instead use the prediction and the estimated value
obtained based on the Kalman filter for the tracked object. This work uses a single video
camera to show how effectively to compute and detect multiple objects concurrently; it is
applied for daytime preprocessing in an automated traffic surveillance system.
Preliminary testing findings show that the suggested algorithm for this vehicle monitoring
system is feasible and effective. It illustrates that using the suggested algorithm with a
single video camera can simultaneously watch, detect, and track several vehicles and
improve execution time. Simulation results on the built system demonstrate that the
proposed system reduced the execution time to approximately 41.5% compared to the
standard background subtraction algorithm. Results indicate the proposed algorithm has
an approximate error for the position and speed of detected and tracked objects compared
with the standard background subtraction algorithm.

Finding an optimal solution to some problem, like minimizing and
maximizing the objective function, is the goal of Single-Objective Optimization (SOP).
Real-world problems, on the other hand, are more complicated and involve a wider
range of objectives, several objectives should be maximized in such problems. No single
solution could be enhanced in all objectives without deteriorating at least one other
goal, which is the definition of Pareto-optimality. Understanding the idea of Multi-
Objective Optimization (MOP) is thus necessary to find the optimum solution. Multi-
objective evolutionary algorithm (MOEA) are made to simultaneously assess many
objectives and find Pareto-optimal solutions, MOEA can resolve multi-objective and
single-objective optimization problems.
This paper aims to introduce a survey study for optimization problem solutions by
comparing techniques, advantages, and disadvantages of SOP and MOP with
metaheuristics and evolutionary algorithms. From this study, we conduct that the
efficiency of MOP lies in the present more than one SOP, but it takes a longer time to
process and train and is not suitable for all applications, While SOP is faster and more
useful in stock and profit maximization applications. And the posterior techniques are
considered the dominant approach to solving multi-objective problems by the use of the
field of metaheuristics.

The article provides an overview of two recent developments in
technology: Business Intelligence (BI) and Deep Learning (DL). In order to
support decision-making processes, BI entails gathering, integrating, and
analyzing data from various sources, while DL uses artificial neural networks
to learn and generate predictions from complicated datasets. This paper
introduces the concepts and principles and highlights recent developments and
applications in different domains of research: education, organizations, stock
market, forecasting, decision-making in real-time, and security. However, the
fundamental problem with the business intelligence approach is that there is no
learning involved. Other limitations and challenges include the capacity that
affects the data analysis process, the variety of data in results, and the need for
a complete presentation of results in the form of dashboards, scorecards,
reports, and portals. The approach choice hinges on the problem\'s context and
requirements and the nature and characteristics of the data. Although BI and
DL are widespread, alternative methods may suit well too, such as machine
learning, data mining, and statistical analysis. Justifying the selection based on
precise needs and goals is crucial. Recurrent neural networks (RNN),
convolutional neural networks (CNN), long short-term memory (LSTM), gated
recurrent units (GRU), and Business intelligence tools are used in the research
problem to address these limitations and explore the potential advantages and
difficulties of integrating BI and DL to achieve an advantage in a given sector.

To perform fault tolerance for Anti-lock Braking System (ABS), This paper
proposes a hybrid Fault Detection and Fault Tolerant Control (FD-FTC) for ABS speed
sensors. It utilizes a Fault Detection (FD) unit and a Data Construction (DC) unit. The
first one, the FD unit, is based on a kNN classifier model with 99.9% fault detection
accuracy to perform three tasks: early fault detection, fault location diagnosis, and
excluding faulty signals from being utilized in further processes. On the other hand, the
second one, the DC Unit, is based on two separate neural network models. These models
have an MSE of 2.01139e-1 and a R2 of 999880 for the first model and an MSE of
1.12486e-0 and 0.999586 for the second model. They are employed to provide an estimated
alternative signal for the ABS speed sensors. These estimated signals are employed to
perform two tasks: confirming fault detection declared by the FD model and compensating
for the excluded faulty signal to fulfill fault accommodation. Both methods are trained and
tested with MATLAB and Simulink. Results demonstrate that the proposed hybrid method
has the ability to accurately detect and tolerate sensor faults and fulfill its design purpose,
especially during emergency braking.

The performance of the Internet is significantly impacted by network
congestion. Because of the internet's current rapid growth, congestion could
increase and cause more packets to be dropped. The Transmission Control
Protocol (TCP) connection is used as the reliable transmission of packets which
has a Drop Tail (DT) mechanism since congestion DT signals occur only when
the queue has become full. The Random Early Detection algorithm mechanism
can be developed as congestion control which eliminates the drawback of a tail
drop queue full. The Red algorithm also has problems dealing with different
numbers of connections due to fix parameter tuning. In this study, the
EXhaustive search used with the RED algorithm to develop the EX-RED
algorithm. Based on network performance metrics such as packet drop, delay,
and throughput, the developed algorithm adjusts the default RED parameter to
find the best of them. When the number of TCP connections changes, the
exhaustive search used systematically enumerates all possible states of
the parameter. The simulation results showed that the EXRED improved
performance of the network as compared with the other five algorithms (GRED,
RED, ARED, NLRED, and NLGRED) by decreasing delay and dropped
packets and increasing packet throughput.

In recent years, safe lockers have been spread in public places to
secure valuable belongings. The people are concerned about losing safe locker
keys or use of the spare key by others and will remain worried about their
things. To solve the forenamed matter, this paper proposed a system that
depends on biometrics to secure valuable things and can minimize their
concerns and worries. The proposed system consists of two major parts,
software, and hardware. In the hardware part, A microcontroller with a camera
and an electronic lock will be used to securely open or close the door of the safe
locker. In the software part, the images that are pictured by the camera will be
prepared by an image processing algorithm and then Support Vector Machines
(SVM) will be trained in the images of a person. The images and information of
the person will be saved until the belongings come out and after that deletes
everything to prepare for another one.