Scribd
Upload
29 views

Course 1

Uploaded by

Tuan Anh Tran
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
29 views

Course 1

Uploaded by

Tuan Anh Tran
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Course I Exam Preparation

Trinh Tuan Phong


November 19, 2024

1 Sentiment Analysis
Exercise 1 Very first mathematical formulation: Linguistic intuition leads to
the Logistic Regression for Sentiment Analysis Classification.
• Derive the gradient descent formula in Logistic Regression case.
Exercise 2 Derive Naive Bayes formulation for Sentiment Analysis Application.

• Maximum Likelihood Estimation.


Exercise 3 - Coding Understand the jupyter notebook of Sentiment Analysis
implementation. You will see some questions like what does the following
function (process_tweet) do?

2 N-gram model
Exercise 1 Mathematical formulation for n-gram models. For example, how to
determine the probability of observing a sequence of words {𝑤 1 , 𝑤2 , . . . , 𝑤 𝑛 }
in bi-gram model.
Exercise 2 Consider the following corpus 𝑉 built on two letters {𝑎, 𝑏}

<𝑠>𝑎 𝑏
<𝑠>𝑎 𝑎
<𝑠>𝑏 𝑎
<𝑠>𝑏 𝑏

Prove that the probability of all possible 𝑘 word sentences over the alpha-
bet , is equal to 1 for any 2.
Exercise 3 𝑘 − 𝑠𝑚𝑜𝑜𝑡 ℎ𝑖𝑛 𝑔 technique formula and its usage.

1
3 Vector embedding
Exercise 1 Vector embedding concept. Static and contextual embedding.
Exercise 2 Understand the concept of occurrence matrix: word by doc design
and word by word design.
• td-idf formula.
• pointwise mutual information formula. Relate it to Information Gain
that you learnt in Machine Learning course.
• Entropy, joint entropy, conditional entropy. Prove various properties
of entropies in the course.
Exercise 3 For word2vec part, we will only focus on Bag-of-Word algorithm.
• Definition of softmax function. How to compute the partial deriva-
tives of this function.
• Two hidden layer Neural Network architecture for Bag-of-Word model.
How to derive the partial derivatives for Gradient Descent update.
Exercise 4 - Coding Understand the Jupyter notebooks on Bag-of-Word imple-
mentation, you will see some questions in the exam like this
What does the following code do?
def forward_prop ( x , W1, W2, b1 , b2 ) :
h = np . dot (W1, x )+ b1
h = np . maximum( 0 , h )
z = np . dot (W2, h)+ b2

return z , h

4 Part-of-Speech (POS) and Name Entity Recogni-


tion (NER)
Exercise 1 Describe Hidden Markov Model (HMM)
• Markov chain
• transition matrix
• emission matrix

Exercise 2 Understand POS problem; BIO, IO labelling.

You might also like