Please enable JavaScript.

Coggle requires JavaScript to display documents.

How the 3 Methods Work, What is Decision Tree, When to Use Decision Tree,…

- - - - Data Classification
      - Regression
        
        The algorithm starts from the root node of the tree
        
        Compares the values of a root attribute with the record (real dataset) attribute and, based on the comparison, follows the branch and jumps to the next node.
        
        For the next node, the algorithm again compares the attribute value with the other sub-nodes and moves further. It continues the process until it reaches the leaf node of the tree.
    - - Prediction Analysis
        
        Use any clustering algorithm that is adequate for the data
        
        Assume the resulting cluster are classes
        
        Train a decision tree on the clusters
- - - - Advantage:
        Simple
        Disadvantage:
        Need large dataset
    - - k-fold, Leave-one-out and stratified
    - - Bagging: averaging the prediction over a collection of classifiers
      - Boosting: weighted vote with a collection of classifiers
      - Ensemble: combining a set of heterogeneous classifiers (Random Forest)
  - - - F and Fß measures
        F measure (F1 or F-score): harmonic mean of precision and recall
        F= (2 x precision x recall)/(precision + recall)
        
        Fß:
        weighted measure of precision and recall
        assigns ß times as much weight to recall as to precision
      - Confusion Matrix
        tool for analyzing how well your classifier can recognize tuples of different classes.
      - Sensitivity and Specificity
        used when significant majority of the negative class and minority of the positive class
        Sensitivity: True Positive recognition rate,
        𝑠𝑒𝑛𝑡𝑖𝑡𝑖𝑣𝑖𝑡𝑦= 𝑇𝑃/𝑃
        Specificity: True Negative recognition rate
        𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐𝑖𝑡𝑦= 𝑇𝑁/𝑁
      - Precision (exactness) & Recall (completeness)
        𝑝𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛= 𝑇𝑃/(𝑇𝑃+𝐹𝑃) = 𝑇𝑃/𝑃′
        what % of tuples that the classifier labeled as positive are actually positive?
        𝑟𝑒𝑐𝑎𝑙𝑙= 𝑇𝑃/(𝑇𝑃+𝐹𝑁) = 𝑇𝑃/𝑃
        what % of positive tuples did the classifier label as positive?
      - Accuracy & Error Rate
        can be applied to most generic problems but is not very useful when it comes to unbalanced datasets
        Accuracy= (TP+TN)/(TP+TN+FP+FN)
        Error rate = (FP+FN)/(TP+TN+FP+FN)
- - - - A collection of data objects
        
        dissimilar (or unrelated) to the objects
        in other groups
        
        similar (or related) to one another
        within the same group
    - - Same group
        
        Instra- cluster distance are minimized
      - Different group
        
        Instra-cluster distances are maximized
  - - - Predictive analytics
        
        To identify future probabilities and trends.
      - Descriptive analytics
        
        To provide information about happened.
    - - Partitional Clustering
        
        k-means Algorithm
        
        k- modes
      - Hierarchical clustering
        
        Agglomerative
        
        Single linkage / MIN
        
        Complete linkage / MAX
        
        Divisive
  - - - As a stand alone tool to get insight into data
        distribution
      - As a preprocessing step for other algorithms
        
        Regression,
        
        PCA
        
        Classification
        
        Association
  - - - market research and customer segmentation
      - biological data and medical imaging
      - search result clustering
      - recommendation engine
      - pattern recognition
      - social network analysis
      - mage processing