Characteristics of Algorithms

Decision Trees Algorithm

K Nearest Neighbor

Algorithm

Naïve Bayes Algorithm

Space Vector machine Algorithm

Random Forest Algorithm

Neural Network Algorithm

Accuracy

The Decision trees have average or moderate level of accuracy

The K nearest algorithm (KNN) also has average accuracy level.

The Naïve Bayes Algorithm has below average and a very poor accuracy level in general.

The space vector machine algorithm has the best overall accuracy.

The random forest algorithm has the highest level of accuracy in general.

The neural network based algorithms also have a fairly moderate level of accuracy.

Type of problem it deals with

The decision trees has the capability of dealing and solving both the classification and as well as regression problems

The KNN algorithm can also deal with either classification or regression problems

The Naïve Bayes Algorithm can specifically deal with only problems of classification.

The space vector machine algorithm is fundamentally used for solving both the classification and regression problems.

The Random forest algorithm can even handle both the

classification and the regression problem

The Neural network algorithm can handle either classification or regression problem

Is it easy to understand the algorithm?

The decision trees are neither easy nor difficult to understand and explain to others. The complexity of a decision tree depends on the complexity of a problem

It is relatively easy to understand a KNN algorithm.

The Naive Bayes Algorithm just like decision trees is neither easy nor difficult to understand.

The space vector machine algorithm is very easy to learn, understand and implement.

The random forest algorithm is very tricky and difficult to understand.

The neural network algorithm is also quite difficult to understand.

Predicting output speed of Algorithm

It does not take much time to analyze data and predicts.

The prediction speed in KNN algorithm depends on the value of

‘n’.

This algorithm also does not take much time for making predictions.

It performs predictions quickly and does not take much time for analyzing data.

The prediction speed of random forest algorithm is relatively moderate, neither too slow nor too fast

The prediction speed of neural network is fast

the capability of handle noise in data

They can moderately tolerate data noise

The KNN algorithm has very low tolerance to noise data

It has exceptional capability of tolerating noise in data

They can moderately tolerate noise in data

They can tolerate noise very well up to certain ration

They can moderately tolerate noise in data

Capability of handling missing values in given data

Its capability of handling missing values in data is relatively good

It cannot handle missing values in data

It can efficiently handle missing values in the data

Its capability of handling missing values in data is average

Its capability of handling missing values in data is relatively good

It cannot handle missing values in data

Ability to automatically learn from experience

They have the ability to learn automatically from past experiences

No, they cannot learn automatically.

No, they also cannot learn automatically.

Yes, it can.

Yes, they have the ability to learn automatically from past experiences

They can learn automatically from past experiences

Interpretability

Moderately fare

It has very low interpretability.

Low Interpretability

Low

Interpretability

High level of interpretability

Low

Interpretability

Performance capability based on little data observation

No, decision trees cannot perform well on the bases of limited data observation

It cannot perform well.

Yes, it can perform very well with only small no of observation.

Yes it can perform well

No, it cannot perform well

No, it cannot perform well