Page 13 - Big Data Analytics for Intelligent Healthcare Management

1.2 BIG DATA ANALYTICAL MODEL 3

Model
Report
Txn
data
Testing Model
Report data validation
generator
Data
sampling
Data Training Data Draft
Clean model
clustering data partition
data
Feature
extraction
Model
Training
Dimension generation
data
aggregation
Data
cubes

FIG. 1.3
Big data analytical model.

Big data process

Data management Analytics

Acquisition and Extraction and Integration and Modeling and Data
recording cleaning aggregation analysis interpretation
FIG. 1.4
Big data process.

which required information can be extracted using data mining. Initially, different types of data come
from different users or devices and the process of data cleansing is performed to remove the irrelevant
data and stores the clean data in the database [8]. Further, data aggregation is performed to store the data
in an efficient manner because incoming data contains a variety of data and a report is generated for
easy use in future. The aggregated data is further stored in data cubes using large storage devices. For
deep analysis, feature extraction is performed using data sampling, which generates the required type of
data. The deep analysis includes data visualization, model learning (e.g., K-nearest-neighbor, Linear
regression), and model evaluation [9].
Fig. 1.4 shows the process of big data, which has two main components: data management
and analytics. There are five different stages in processing big data: (1) acquisition and recording