Research ProposalPDF Available

Machine Learning Algorithms

Authors:
Vansh Jatana at SRM Institute of Science and Technology
Vansh Jatana
Download full-text PDF
Read full-text
Download citation

Abstract

Machine learning is a branch of artificial intelligence that allows computer systems to learn directly from examples, data, and experience.It has many algorithms and unfortunately we are unable to select the right algorithm for the right problem.Following are the factors which help us to choose the correct algorithm Factors help to choose algorithm 1-Type of algorithm 2-Parameterization 3-Memory size 4-Overfitting tendency 5-Time of learning 6-Time of predicting
Content uploaded by Vansh Jatana
Author content
All content in this area was uploaded by Vansh Jatana on May 07, 2019
Content may be subject to copyright.
Machine Learning Algorithms
Machine learning is a branch of artificial intelligence that allows computer systems to learn
directly from examples, data, and experience.It has many algorithms and unfortunately we are
unable to select the right algorithm for the right problem.Following are the factors which help us
to choose the correct algorithm
Factors help to choose algorithm
1-Type of algorithm
2-Parametrization
3-Memory size
4-Overfitting tendency
5-Time of learning
6-Time of predicting
Type of algorithm
1. Regression
It is a technique used to predict dependent variable in set of independent variable.The
algorithms which come under regression are
1-Linear Regression
2-Decision Tree
3-Random Forest
4-Boosting
2 . Classification
It is a technique used for approximating a mapping function (f) from input variables (X) to
discrete output variables (y).The algorithms which come under classification are
1-Logistic Regression
2-Naive Bayes
3-SVM
4-Neural Networks
5-Decision Tree
6-Random Forest
6-Boosting
3. Clustering
It is a technique for dividing the population or data points into a number of groups such
that data points in the same groups are more similar to other data points in the same
group and dissimilar to the data points in other groups. K-means is important algorithm
used for clustering
Note
Decision tree, Random forest and Boosting are the algorithm which can be used for both
classification and regression
Parametrization
Parameters are key to machine learning algorithms. They are the part of the model that is
learned from historical training data.Parameters are key to machine learning algorithms. They
are the part of the model that is learned from historical training data.We are classifying our
parameters as
1-No parameters
2-Weak
3-Simple/Intuitive
4-Not Intuitive
Memory Size
It is the space we need to store our data and variables.Researchers are struggling with the
limited memory bandwidth of the DRAM devices that have to be used by today's systems to
store the huge amounts of weights and activations in DNNs.GPUs and other machines
designed for matrix algebra also suffer another memory multiplier on either the weights and
activations of a neural network. We are classifying our memory size required as
1-Small
2-Large
3-Very Large
Overfitting Tendency
When a model tries to predict a trend in data that is too noisy. Overfitting is the result of an
overly complex model with too many parameters. A model that is overfitted is inaccurate
because the trend does not reflect the reality of the data.There are many techniques that can be
use to mitigate overfitting, including cross-validation, regularization, early stopping, pruning,
Bayesian priors, dropout and model comparison.
We are classifying Overfitting tendency as
1-Low
2-Average
3-High
4-Very high
Time for Learning
Time for learning is time associate with training of dataset.It varies with size of data and
algorithm we are using in that. We are classifying time for learning as
1-Weak
2-Costly
3-Very Costly
Time for Predicting
Time of predicting is time associate with testing of dataset.It varies with size of data and
algorithm we are using in that.We are classifying time for learning as
1-Weak
2-Costly
Algo
Type
Parametrization
Memory Size
Overfitting
Tendency
Time for
Learning
Time for
Predicting
Linear
Regression
R
Weak
Small
Low
Weak
Weak
Logistic
Regression
C
Simple
Small
Low
Weak
Weak
Decision
Tree
R & C
Simple/ intuitive
Large
Very High
Weak
Weak
Random
Forest
R & C
Simple/ intuitive
Very Large
Average
Costly
Costly
Boosting
R & C
Simple/ intuitive
Very Large
Average
Costly
Weak
Naive
Bayes
C
No parameters
Small
Low
Weak
Weak
SVM
C
Not intuitive
Small
Average
Costly
Weak
Neural
Networks
C
Not intuitive
Inter
Average
Costly
Weak
K-Means
CL
Simple/ intuitive
Large
High
Weak
Weak
... shows that an object set can be decomposed into many sub-object sets by decision attribute information. Furthermore, according to category representation axioms [54], we have every reason to believe that different subconcept spaces where each sub-concept space consists of various concepts can be constructed by different class labels in a label set Y. Hence, hereinafter, we discuss the situation of a sub-concept space that is associated with a single class label under a regular formal decision context. ...
... According to sample separation axiom [54], for any instance, there always exists a unique class that is most similar to it. Hence, given an instance x, the class vector can be generated as follows: each sub-concept space will first produce a set of CS degrees by computing the CS degree between the given concept and any concept from a sub-concept space. ...
... , n}) and S(C i , G αr,k i−1 ) = Sim(C i , G αr,k Fi−1,Hi−1 ). Furthermore, inspired by [54], the category similarity function between the given concept C i and a class space G αr,k Fi−1,Hi−1 can be defined as ...
View
... Machine learning: is a branch of artificial intelligence that allows computer systems to learn directly from examples data, and experience [6]. Machine learning is the science of getting computers to learn and act like humans do, and improve their learning over times. ...
A Machine Learning Model for Malware Detection Using Recursive Feature Elimination (RFE) For Feature Selection and Ensemble Technique
Article
Full-text available
  • Aug 2022
In recent years, almost every member of the society has been using Internet for their daily life. This is because it has become an integral part of our life. But, as the technology is continually advancing, the cyber security threats such as spamming and malwares, are swarming cyberspace exploiting vulnerabilities, making it harder for computer users to keep their data safe. To be able to keep people and files safe, an antivirus programs were created to aid in malware detection and prevention. While the performance of antivirus is commendable, they suffer from two major drawbacks, their inability to detect new malwares whose signatures are not available in the virus definition at the time, and its inability to detect zero day attack. Recently, researchers proposed machine learning models that are capable of detecting zero day attacks, but these models are also vulnerable to an adversarial attacks. To overcome this problem, an ensemble boosting model for malware detection is proposed in this research, and also the role of Recursive Feature Elimination (RFE) algorithm on PE header files dataset is investigated. The proposed model achieved a state of the art accuracy of 99%, precision of 98% and Recall of 97%. The model also outperformed other existing model when their performances were compared.
View
... For a given repetitive motion of the robot, a set of positions of the end-effector can be gathered in a first stage to mathematically model the trajectory of the programmed movement. Based on this model considered as a reference model, the trend of the accuracy errors can be measured by calculating the residuals between new positions and the model defined Data cleaning operations are performed to remove the non-relevant records captured right before the robot starts moving or right after it stops and only keep the needed values when the robot is performing the repetitive motion [4]. Once the data is cleaned, the power and positions data sets must be segmented into multiple sub-sets where each of them corresponds to some similar patterns related to the repetitive movement of the robot. ...
Predictive Maintenance Of Industrial Robots
Conference Paper
Full-text available
  • Aug 2020
The objective of this project is predictive maintenance of industrial robots and the possibility of building a condition monitoring system based on the data analysis of robot’s power measurements and detecting the robot accuracy error based on electricity data analysis for predictive maintenance.An experimental procedure is carried out to oversee the correlation between the robot accuracy error and a set of extracted features from current time-series, and to evaluate the proposed predictive modeling
View
... , 1}. 2) Comparison with Similarity-Based Classification Methods: Since the proposed method is implemented based on the fuzzy conceptual clustering which was completed by the fuzzy concept similarity, we compare it with several similarity-based classification methods. According to the fuzziness of classifiers, these methods were classified into two categories: (1) two popular fuzzy nearest neighbor algorithms, namely IF-KNN [54] and FENN [54], and five state-of-the-art similarity-based fuzzy classification methods [54], that is, PFKNN, CFKNN, FRNN, FRNN-FRS and FRNN-VQRS; (2) the standard K-Nearest Neighbours (K-NN) [19], several popular similaritybased classification methods such as Radius Neighbors Classifier (R-NN) 5 , KD-Tree [55] and Ball-Tree [56], and a state-ofthe-art CCL model (CCLM) [13] which can also be considered as a method with the concept similarity. The first category is acknowledged as fuzzy classifiers in which fuzzy measures were introduced to deal with fuzziness and ambiguity of data, and the second category is known as crisp classifiers that all the characteristics of data can be described accurately. ...
View
... Whereas machine learning is a method of data analysis not relying on specific instructions but learning independently from the data. Machine learning algorithms [7] of varying complexity are used in the analysis of big data. For example, convolutional neural networks are used for deep learning and principal component analysis for dimensionality reduction. ...
View
... we know that CCL was studied from a cognitive perspective by many scholars [8,12,13,11,17,20,41,42], and to meet of different requirements, various of names were used such as cognitive model [41], cognitive system [34] and cognitive concept learning [12,42]; however, in order to stress the importance of the concept in machine learning and cognitive process, its name was also discussed by Mi and Li in 2016; and then we first used the concept of concept-cognitive learning in 2016 [17] and it is called classical CCL; moreover, an atomic concept and a composite concept were also defined in [17]. Based on it, inspired by the similarity degree [38] and a formal decision context [40], Mi et. al began to explore how to incorporate it into machine learning in 2016, and then a novel CCLM was proposed in 2018 [24]. ...
A generalized concept-cognitive learning: A machine learning viewpoint
Article
Full-text available
  • Jan 2018
Concept-cognitive learning (CCL) is a hot topic in recent years, and it has attracted much attention from the communities of formal concept analysis, granular computing and cognitive computing. However, the relationship among cognitive computing (CC), concept-cognitive computing (CCC), CCL and concept-cognitive learning model (CCLM) is not clearly described. To this end, we first explain the relationship of CC, CCC, CCL and CCLM. Then, we propose a generalized concept-cognitive learning (GCCL) from the point of view of machine learning. Finally, experiments on some data sets are conducted to verify the feasibility of concept formation and concept-cognitive process of GCCL.
View
View
View
ResearchGate has not been able to resolve any references for this publication.