IoTDataMiningFramework

---

Usage

1. Generate IoTDataMiningFramework.jar file.
2. Deploy and import IoTDataMiningFramework.jar file in eclipse project.
3. Make sure you have the right pom file with dependencies.
4. Extend the existing models to create specialized ones

---

API

Interface DataResource

Package: Framework.resource

Modifier and Type	Method and Description
void	closeFile()
void	getLine(String filename, String delimiter, int num)
void	openFile(String filename)
void	readAll(String filename, String delimiter)
ArrayList of String	readData(ArrayList of DataPoint data, String delimiter, int[] columns, int lblCol, DataPointType type)
ArrayList of DataPoint	readData(String delimiter, int column, DataPointType type)
void	readSome(String filename, String delimiter, int limit)

Abstract Class DataPoint

Package: Framework.data.DataPoint

Modifier and Type	Method and Description
abstract Label	getLabel() Assigned label of training data.
String	getName()
int	getSequenceID()
java.sql.Timestamp	getTimestamp()
DataPointType	getType()
java.lang.Object	getValue()
abstract java.lang.Object	getValue(java.lang.String column) Get sample data value from specified column.
boolean	has(Feature feature)
void	setName(java.lang.String name)
void	setSequenceID(int seq)
void	setTimestamp(java.sql.Timestamp timestamp)
void	setType(DataPointType type)

Interface TimeSeries

Package: Framework.algorithm.Timeseries

Modifier and Type	Method and Description
double	addDataPoint(T dp)
void	calculateStdDev(T dp)
double	getAverage()
double	getStdDev()
void	reset()

Abstract Class DecisionTree

Package: Framework.algorithm.classification.decisiontree.DecisionTree

Modifier and Type	Method and Description
Label	classify(DataPoint dataSample) Classify dataSample.
Node	getRoot() Get root.
void	printSubtree(Node node)
void	printTree()
void	train(List of DataPoint trainingData, List of Feature features) Trains tree on training data for provided

Interface ImpurityCalculationMethod

Package: Framework.algorithm.classification.decisiontree.impurity

Modifier and Type	Method and Description
double	calculateImpurity(List of DataPoint splitData) Calculates impurity value.
default	double getEmpiricalProbability(List of DataPoint splitData, Label positive, Label negative) Calculate and return empirical probability of positive class.

Abstract Class Label

Package: Framework.algorithm.classification.decisiontree.label

Modifier and Type	Method and Description
abstract boolean	equals(Object o) Force overriding equals.
abstract String	getName()
abstract String	getPrintValue() Label value used to print to predictions output.
abstract int	hashCode() Force overriding hashCode.