Core concepts

This chapter gathers HephIA’s concepts glossary.

Along your data crunching journeys using HephIA product you will need to manipulate many concepts, some of them have already been established by the community when others are specific to us.

Some of these concepts exist on both side, they often share the same semantic. We simply add the little something that allow you to dig down into your problematics. To make it easier for you will use following syntactic rules to distinguish :

• Well established concepts on both theoretical and / or technical sides used Roman style.
• HephIA own concepts are in italic.
• Concepts merging well established and HephIA ones are in bold.

Page organization :

• Concepts
• Minimal required concepts

Concepts

• Id
• Model
• Instance
• Predictor // is there exists a general concept of predictor, i do not believe it, but check and/or ask others
• Evaluator // is there exists a general concept of predictor, i do not believe it, but check and/or ask others
• Evaluation // is there exists a general concept of predictor, i do not believe it, but check and/or ask others
• Pipe // is there exists a general concept of predictor, i do not believe it, but check and/or ask others
• Algorithm
  • Clustering
    • Hard clustering
    • Soft clustering
  • Dimensions reduction
  • Outliers detection
  • Quality indices
    • Internal
    • External
  • Vectorization
    • Categorical binarization // It is properly define theoretically but done technically by us, so which syntax
    • Image
    • Text
    • Audio
    • Video
    • Monovariate time series
    • Multivariate time series
• Scalability
  • Algorithmic complexity
    • Time
    • Space
• DAG : Directed Acyclic Graph

Minimal required concepts

The minimal required concepts’set include for few various examples :

• Clustering

  • Cluster
  • Dissimilarity measure
  • ClusterId

• Dimensions reduction

• Outliers detection

• Predictor

• At least algorithms you want to apply

Quick overlook of some important concepts

1. Model
2. TraversableModel
   1. TraversableMemory
3. GlobalModel
   1. GlobalMemory
4. TreeModel
   1. TreeMemory
   2. GTreeMemory
   3. GTreeModel
5. Instance
   1. InstanceMemory
   2. List, Vector, Array, ParVector, ParArray, RDD InstanceMemory/Instance
   3. GInstanceMemory
   4. GInstance
   5. List, Vector, Array, ParVector, ParArray, RDD InstanceMemory/Instance
6. Extension
   1. GExtension
7. Clustering
   1. ClusteringEntity
      1. Discrete Distribution
   2. Hard Clustering
      1. HardClusteringEntity
         1. Constant
      2. Aggregator
   3. Soft Clustering
      1. Categorical Clustering
         1. CategoricalDistribution
8. Cluster
   1. GInstanceMemory
   2. ProtoCluster
9. Clusters
   1. ProtoClusters
10. Pipe
11. Algorithm1
    1. TraversableAlgorithm
       1. GTraversableAlgorithm
    2. InstanceAlgorithm
    3. …
    4. ClusteringAlgorithm
    5. ClustersAlgorithm
    6. KMeans
    7. IterativeAlgo
12. Predictor
13. Evaluator
    1. Evaluation
    2. Internal / External
14. FuncData
    1. FuncDataExtension
15. Outlier detection
16. Dimension reduction
17. Matrix
18. NN
    1. KNN
    2. ANN
       1. Hash
19. Preprocessing
    1. Gradient ascent
    2. Vectorizer