Advanced

Computer
Organization and
Architecture
ASSIGNMENT

Name: Pranjal Kalita

Roll no: CS-28
MSC 1st Sem
 Machine Learning
Machine Learning is an exciting branch of Artificial Intelligence, and it’s all around us. Machine
learning brings out the power of data in new ways, such as Facebook suggesting articles in your feed.
This amazing technology helps computer systems learn and improve from experience by developing
computer programs that can automatically access data and perform tasks via predictions and
detections.

As you input more data into a machine, this helps the algorithms teach the computer, thus
improving the delivered results. When you ask Alexa to play your favourite music station on Amazon
Echo, she will go to the station you played most often. You can further improve and refine your
listening experience by telling Alexa to skip songs, adjust the volume, and many more possible
commands. Machine Learning and the rapid advance of Artificial Intelligence makes this all possible.

Importance of Artificial intelligence

1. Studying AI and Machine Learning promises a bright career.

With every company looking to introduce Artificial Intelligence in their domain, studying AI
and Machine Learning opens up a world of opportunities to create cutting-edge technologies in
diverse sectors, such as computer protection, image recognition, pharmacy, or face recognition.

2. Artificial Intelligence and Machine Learning improves user experience.

AI does not necessitate the use of a different app or device. It enhances the intelligence of the
services we use on a daily basis.

A mix of AI technology, such as chatbots, emulation, and virtual assistants like Google
Assistant, is helping to increase user experience by incorporating many helpful features to an existing
app.

3. Benefits of machine learning in Heath care system:

o Improving diagnosis.
o Developing new treatments / drug discovery / clinical trials.
o Reducing costs:

Machine learning technologies can be used by healthcare organizations to improve the efficiency
of healthcare, which could lead to cost savings. For example, machine learning in healthcare could be
used to develop better algorithms for managing patient records or scheduling appointments. This type
of machine learning could potentially help to reduce the amount of time and resources that are wasted
on repetitive tasks in the healthcare system.

4. Machine learning is essential for self-driving cars

While many are rightfully wary of self-driving cars right now, these cars will become more
common. The secret is machine learning. The algorithms collect data via sensors and cameras, analyse
the data, and decide what the car should do. One team at Boston University recently created a “watch
and learn” algorithm that taught self-driving cars to drive by watching other cars. In a test
set in two virtual towns, the self-driving neural networks got into very few accidents and reached their
destinations 92% of the time. Studies like this show the potential of machine learning in self- driving
cars.

5. Machine learning could take over dangerous jobs

Many jobs put human life at risk. Nuclear cleanup is a big one. In 2021, scientists participated in a
consortium focused on using AI and robotics in nuclear environments. At Chernobyl, arguably the most
famous nuclear site, the team trained robots to create a 3D map and measure radiation. Using
machine learning, robots can be trained to recognize the differences between radioactive waste types.
This would help humans safely identify and get rid of nuclear waste. Machine learning could also make
robots very effective at jobs involving dangerous chemicals, extreme heavy lifting, and fires.

Brief history of Machine Learning time line

Brief description about Machine Learning history:

Early Foundations (1940s-1950s):

1. McCulloch-Pitts Neuron (1943): The concept of an artificial neuron, the basic building block
of neural networks, was introduced by Warren McCulloch and Walter Pitts. This laid the
groundwork for later neural network developments.

2. Turing’s Test (1950): Alan Turing proposed the Turing Test, a benchmark for determining a
machine’s ability to exhibit intelligent behavior indistinguishable from that of a human. While
not directly related to ML, this idea sparked discussions on machine intelligence
Birth of Machine Learning (1950s-1960s):

1. Rosenblatt’s Perceptron (1957): Frank Rosenblatt developed the perceptron, an early form of
a neural network. The perceptron could learn to distinguish between patterns, laying the
foundation for neural network research.

2. Dartmouth Conference (1956): The Dartmouth Conference marked the birth of artificial
intelligence (AI) as a field. Early AI pioneers, including John McCarthy and Marvin Minsky,
aimed to develop machines capable of learning from experience.

Winter of AI (1970s-1980s):

1. Expert Systems (1970s-1980s): AI research shifted towards expert systems — rule-based

systems designed to mimic human expertise. While successful in certain applications, they
struggled with handling uncertainty and lacked adaptability.

Renaissance of Machine Learning (1980s-1990s):

1. Backpropagation Algorithm (1986): The development of the backpropagation algorithm by

Geoffrey Hinton and colleagues allowed neural networks to efficiently adjust their weights
during training, overcoming limitations of earlier models.

2. Support Vector Machines (1990s): Vladimir Vapnik and Corinna Cortes introduced support
vector machines, a powerful algorithm for classification and regression tasks, contributing to
the diversification of ML techniques.

Rise of Big Data and Computational Power (2000s):

1. Big Data Era: The proliferation of digital data, along with advances in data storage and
processing, provided the fuel for ML algorithms. The ability to handle vast datasets became a
defining feature of successful ML applications.

2. Ensemble Learning (2000s): Techniques like boosting and random forests gained popularity,
demonstrating the power of combining multiple models to improve overall performance and
robustness.

Deep Learning Resurgence (2010s-2020s):

1. ImageNet and Deep Learning (2012): The ImageNet competition saw the rise of deep learning,
particularly convolutional neural networks (CNNs). AlexNet’s victory marked a paradigm shift,
leading to breakthroughs in image and speech recognition.

2. Reinforcement Learning (2013–2015): DeepMind’s AlphaGo showcased the potential of

reinforcement learning, combining neural networks with reinforcement learning techniques
to achieve superhuman performance in complex games.

3. Transformer Architecture (2017): The introduction of the Transformer architecture,

particularly through models like BERT (Bidirectional Encoder Representations from
Transformers), revolutionized natural language processing tasks.
 Current Landscape:

1. AutoML and Transfer Learning: Recent years have seen the rise of AutoML, making machine
learning more accessible to non-experts. Transfer learning techniques enable models to
leverage pre-trained knowledge for new tasks.

2. Ethical Considerations: The ML community is increasingly addressing ethical concerns,

including bias and fairness in algorithms, emphasizing the importance of responsible AI
development.

3. OpenAI’s GPT-3 (2020): GPT-3, the third iteration of OpenAI’s Generative Pre-Trained
Transformer, demonstrated unprecedented language generation capabilities, pushing the
boundaries of what AI can achieve.

Application of Machine Learning

1. Image recognition:

One of the most common uses of machine learning is image recognition. To do this, data
professionals train machine learning algorithms on data sets to produce models capable of recognizing
and categorizing certain images. These models are used for a wide range of purposes, including
identifying specific plants, landmarks, and even individuals from photographs.

2. Fraud detection:

Financial institutions process millions of transactions daily. Perhaps unsurprisingly, it can be

difficult for them to know which are legitimate and which are fraudulent.

As more and more people use online banking services and cashless payment methods, the
number of fraudulent transactions has similarly risen. In fact, according to a 2023 report from
TransUnion, the number of digital frauds attempts in the US rose a staggering 122 percent between
2019 and 2022.

3. Personal assistants and voice assistants:

It’s ML that powers the tasks done by virtual personal assistants or voice assistants, such as
Amazon’s Alexa and Apple’s Siri. This communication can involve speech recognition, speech-to-text
conversion, NLP, or text-to-speech. When someone asks a virtual assistant a question, ML searches
for the answer or recalls similar questions the person has asked before.

4. Product recommendations:

Machine learning is widely used by various e-commerce and entertainment companies such as
Amazon, Netflix, etc., for product recommendation to the user. Whenever we search for some
product on Amazon, then we started getting an advertisement for the same product while internet
surfing on the same browser and this is because of machine learning.

5. Self-driving cars: One of the most exciting applications of machine learning is self-driving cars.
Machine learning plays a significant role in self-driving cars. Tesla, the most popular car
manufacturing company is working on self-driving car. It is using unsupervised learning method to
train the car models to detect people and objects while driving.

6. Virtual Personal Assistant:

We have various virtual personal assistants such as Google assistant,

Alexa, Cortana, Siri. As the name suggests, they help us in finding the information using our voice
instruction. These assistants can help us in various ways just by our voice instructions such as Play
music, call someone, open an email, Scheduling an appointment, etc.

7. Sentiment Analysis:

Sentiment analysis is one of the most necessary applications of machine learning. Sentiment
analysis is a real-time machine learning application that determines the emotion or opinion of the
speaker or the writer. For instance, if someone has written a review or email (or any form of a
document), a sentiment analyser will instantly find out the actual thought and tone of the text.
This sentiment analysis application can be used to analyse a review-based website, decision-
making applications, etc.

Types of machine Learning Algorithm

Supervised Machine Learning

Supervised learning is the types of machine learning in which machines are trained using well
"labelled" training data, and on basis of that data, machines predict the output. The labelled data
means some input data is already tagged with the correct output.

In supervised learning, the training data provided to the machines work as the supervisor that teaches
the machines to predict the output correctly. It applies the same concept as a student learns in the
supervision of the teacher.

Unsupervised Learning

Unsupervised learning is a branch of machine learning that deals with unlabelled data. Unlike
supervised learning, where the data is labelled with a specific category or outcome, unsupervised
learning algorithms are tasked with finding patterns and relationships within the data without any
prior knowledge of the data’s meaning. This makes unsupervised learning a powerful tool for
exploratory data analysis, where the goal is to understand the underlying structure of the data.

Reinforcement learning (RL) is an interdisciplinary area of machine learning and optimal control
concerned with how an intelligent agent should take actions in a dynamic environment in order to
maximize a reward signal. Reinforcement learning is one of the three basic machine learning
paradigms, alongside supervised learning and unsupervised learning.

Ensemble Learning : Ensemble learning is a machine learning technique that combines the predictions
from multiple individual models to obtain a better predictive performance than any single model. The
basic idea behind ensemble learning is to leverage the wisdom of the crowd by aggregating the
predictions of multiple models, each of which may have its own strengths and weaknesses. This can
lead to improved performance and generalization.