|
ML algorithm | Advantages | Disadvantages |
|
DT | (i) Very simple and fast | (i) Requires long time to train the model |
(ii) Not affected by the increase of the dimensionality of the data | (ii) Requires larger amount of memory for analyzing large databases |
(iii) The model is easily understood (plausible) | (iii) Not suitable for problems that require diagonal partitioning |
(iv) Generates good accuracy based on the quality of the data | (iv) Can generate a complex representation for some concepts due to the replication |
(v) Support incremental learning | (v) The order of the attributes affects the performance |
|
RF | (i) Can resist overfitting | (i) Difficult to interpret the model |
(ii) Does not require attributes selection | (ii) Correlated variables cause performance degradation |
(iii) Model variances decrease with the increase in the number of trees | (iii) Reliance on the random generator of the implementation |
(iv) Increasing the number of trees does not affect the bias of the model |
|
SVM | (i) One of the most robust and accurate algorithms | (i) Requires intensive computations to build the model |
(ii) Requires few data for training and is not affected by the dimensionality of the data | (ii) Very slow in the learning phase |
(iii) Generate the best function to conduct binary classification | (iii) Memory required doubles as the number of training instances increases |
(iv) Less prone to overfitting compared to other algorithms | (iv) It is difficult to interpret the results of the model |
|
ANN | (i) Tolerable to noise and able to predict new classes | (i) Requires a long time to train a model |
(ii) Applicable even if the relation between attributes and classes is not well defined | (ii) Very difficult to interpret how the model works |
(iii) Suitable for continuous values | (iii) Requires empirical adjustment of different parameters such as the number of hidden layers and the number of nodes in each layer |
(iv) Computations can be accelerated due to ANN parallel nature |
(v) Applied successfully to different real-world issues such as handwritten character recognition and laboratory medicine |
|
NB | (i) Requires short time for training and it is easy to construct the model | (i) Theoretically, classifiers based on NB algorithms have a low error rate. However, in practice, this is not entirely true due to the assumption that different attributes are independent of each other |
(ii) Can be computationally optimized | (ii) Yields low accuracy results compared to other ML algorithms |
(iii) Can be used with large datasets |
(iv) Outcomes are easily interpreted |
(v) It operates in a well and robust manner even though it might not be the best algorithm for a certain application |
|