174

R. O. Ogundokun et al.

Toincreasetheaccuracyofoutcomes,datadependabilityandexchangearecritical.

The use of ML and BT together can produce extremely exact results. As a result,

this study gives a thorough analysis of machine learning acceptance for making

Blockchain-based IoT network smart applications further robust to attacks in this

study. To investigate threats on a blockchain-based IoT network, an improved K-

Nearest Neighbor (KNN) classifier is presented.

Numerous research works happen to address ML procedures for blockchain-based

IoT applications; however, they have not yet been fully explored. In this study, the

authors looked into the use of machine learning for blockchain-based IoT network

smart applications. This paper’s research contribution is listed below:

1.

A quick look at how machine learning, the internet of things, and blockchain

may be used to create smart application architecture.

2.

To investigate threats on a blockchain-based IoT network, an improved

K-Nearest Neighbor (KNN) classifier is presented.

The rest of the manuscript is prearranged as Sect. 2 discussed the material and

method used for the execution of the study proposed classifier. Section 3 discussed

the results discovered and the interpretation of the results. The paper was concluded

in Sect. 4 and future work was also proposed as well.

2

Material and Method

This section discussed the datasets used for the execution of the system. The clas-

sifier employed for the execution and the performance metrics used for the study is

discussed as well.

2.1

Datasets

ThedatasetusedfortheexecutionoftheKNNclassifiersistheIoTblockchaindataset.

The dataset was downloaded from the Mendeley database repository. They consist

of 17 features and 81 instances. The dataset can be found using this link: https://data.

mendeley.com/datasets/rxsdfg8ct9/1. https://doi.org/10.17632/rxsdfg8ct9.1.

Bio studies: Supporting data is http://www.ebi.ac.uk/biostudies/studies/S-EPM

C6412473?xr=true.

2.2

Proposed System

A supervised ML classifier K-Nearest Neighbor (KNN) was used for the system

execution. The IoT blockchain dataset was employed to assess the proficiency and