IOTrust Chain

KNN (K-Nearest Neighbours)

KNN is a supervised learning algorithm that assigns labels to the new datapoint based on its distance from the old data points on which the model was trained. The symbol “K” represents the number of nearest neighbours.

ADVANTAGES:-

1.)Can be used for both regression and classification problem statements.
2.)Can work on Non-Linear Data.
3.)Simple to understand and implement.
4.)Has a single hyperparameter that is K which makes the hyperparameter tuning easy.

DISADVANTAGES:-

1.)Sensitive to outliers.
2.)Doesn’t work well on Large Datasets as calculating distances between each data point will be costly.
3.)Data in all the dimensions should be scaled using feature engineering.
4.)Does not work well on imbalanced data.

Lockbit Again Into the Headlines, with new links between LockBit and BlackMatter/DarkSide - LockBit 3.0/Lockbit Black

The Lockbit Ransomeware group has been in massive spotlight in the IT world for the last 3 months, due to its vigorous nature. From claims of attacks on the likes of the Italian Tax Agency, the whole St. Marys town, to startling announcements of the first Ransomeware Bug Bounty Program with its new 3.0 release.

Now, the new 3.0 variant, launched in June 2022, is being linked with inactive ransomware BlackMatter & DarkSide due to their similarities.
The fact that Multiple portions of LockBit 3.0’s code are borrowed from the BlackMatter ransomware was revealed during the analysis of a sample. LockBit's extensive similarities to BlackMatter come from overlaps in the privilege escalation and harvesting routines used to identify APIs required to terminate processes and other functions as well as the use of anti-debugging and threading techniques designed to thwart analysis.
Rumors are that Lockbit may have hired former BlackMatter developers, whom BlackMatter supposedly fired after a flaw allowing victims to decrypt their files was discovered, justifying its name "Lockbit Black"

Let's Secure the Internet - IOTrust Chain

Self taught learning is a mechanism of machine learning that refers to getting the machine trained from the unlabelled data and let it predict on the previously unseen data.

The training process majorly includes 3 steps:-
1. Learning via high-level representations from unlabelled data
2. Reconstructing labelled data
3. Building classifier over reconstructed data

For example, Self Driving cars or Cars with autopilot mode use self-taught learning technology to learn from the earlier movement of the car. It uses this technology to learn the movement of the car in various situations such as in traffic, at red lights, in tunnels and various other situations.