Selecting the Right Type of Algorithm for Various Applications - Phdassistance
                     An Academic presentation by
Dr. Nancy Agnes, Head, Technical Operations, 
Phdassistance  Group www.phdassistance.com
Email: [email protected]
TODAY'  S
DISCUSSION
Introduction
Understanding the
Data  Required
Accuracy
Speed
Parameters
INTRODUCTION
M achine learning algorithms may be  
classified mainly into three main types.
Supervised learning constructs a  
mathematical model from the training  
data, including input and output labels.
T he techniq ues of data categorization  
and regression are deemed supervised  
learning.
Contd...
In unsupervised learning, the system constructs a model using just the input  
characteristics but no output labeling.
The classifiers are then trained to search the dataset for a specific pattern.  
Examples of uncontrolled learning algorithms including clustering and  
segmentation.
In reinforcement learning, the model learns to complete a task in  reinforcement 
learning by executing a number of actions and choices that it  improves itself 
and then understands from the information from these actions  and decisions 
(Lee & Shin, 2020).
UNDERSTANDING
T HE
DAT A The f i rst and primary stage in determining an  a lg orithm is the understanding of your data.
One needs to acquaint themselves with data  
before thinking about the various algori thms.
One easy approach of doing this is to view  
data and attempt to detect patterns in them,  
to watch their behavior and especial ly their  
size.
Contd...
The size of the data is an important parameter. Some algorithms do better  than 
others with greater data (Mahfouz et al., 2020).
For instance, algorithms with higher bias or lower variance classification are  
more effective than lower bias or higher variance classifications in limited  
training datasets (Richter et al., 2020).
For instance, Naïve Bayes will do better than kNN if the training data is  smaller.
Figure 1: Types o f  Machine Learning 
Algorithms
The feature of data is
parameter . The another  way the
created, and whethe dr ai ttais ils inear
to  the data must be
considered.
Tsuhietendm, aysbucehaasl ineraergmreosdseiol nis or  
m SoVsMt . However , i f youris data 
more compl icated  then more
compl icated algorithms l ike
Random forest may be required.
Contd...
The features being l inked or sequentia l
a lso  requires specif ic type of algori thms.
The type of data is an important parameter  
( Vabalas et al., 2019 ) . The data maybe classif ied  
into input or output.
Use a supervised learning method i f the input  data 
are labeled; otherwise, unsupervised
algorithm must be used. I the output is  
f  numerical , on the other then regression
whailnl db,e used, but i f i t is a col lect ion of groups, i 
t  is an issue of clustering .
Contd...
REQUIRED 
ACIn CtheU neRxt AstCep,Y i t should be decided  
whether or not accuracy is important for  
the issue one is attempting to address .
The accuracy of an application refers to  
the capacity of an individual method to
estimate a response f rom a given  
observat ion near to the r ight
( Garg, 2020 ) response
. Contd..
.
Sometimes a correct reply to
our  target application is not
essential.
I f the approximat ion is strong
enough, by adopting an 
approximate model ,  
 we  reduce t may
acondnspidroecraebslsyingthte ime. raining
Approximat ion approaches, such as
l inear of non- l 
regression  data, inear  do not
pdraetaveonvt eorrf i t t ing . execute
S  P  E  E  
D
Sometimes users have to choose  between 
speed and accuracy in order to  decide on 
an algori thm.
Typical ly, more precision takes longer to  
achieve, over a longer t imeline, while  
faster processing has less accuracy.
The incredibly simple algori thms  l ike  
Naïve Bayes and Logistic regression are  
used often since they' re simple, quick to  
run algori thms.
C o n td . . .
Using more advanced techniques l ike support vector machine  l 
earning, neural networks, and random forests, might take a lot  
longer to learn, and would also give higher accuracy.
Therefore, the question is how much is the project worth, Is  t ime 
more important or the accuracy .
I f i t is t ime, simpler methods must be used, while i f accuracy is  
more important, then one has to go with more sophist icated  ones.
P  A  R  A  M  E  T  E  R  
S
The parameters will impact how the  
algorithm behaves . Options that alter the  
algori thm' s behavior, such as tolerance  
for error or the number of i terat ions.
For as many parameters as the data has,  t 
ime required to process the data t raining  
and processing t ime is f requently  
proport ional.
C o n td .. .
The the number of parametersthe  model' 
greater  the more t i t takes to process s  and t 
dimensions ,  algori thm ime  numerous parameters rain.  
 mHeotwheovde irs, adnaptable. with means the
M achine learning addresses measurable variables. Having more  
features might slow down certain algori thms, therefore this  
causes them to take a lengthy t ime to t rain.
So long as the issue has a large feature set, one should choose  
an algori thm such as SVM, which is best suited to those with  
numerous features.
C O N T A C T  U S
UNITED KINGDOM
+44 7537144372
INDIA
+91-9176966446
EMAIL
[email protected]