CTF

การแข่งขัน CTF (Capture The Flag) ในบริบทความปลอดภัยทางไซเบอร์ คือเกมจำลองสถานการณ์เพื่อทดสอบและพัฒนาทักษะด้านความปลอดภัยของคอมพิวเตอร์ โดยผู้เข้าแข่งขันจะต้องค้นหา "ธง" (Flag) ที่ซ่อนอยู่ผ่านการแฮ็ก เจาะระบบ หรือแก้โจทย์รูปแบบต่างๆ

QKD vs. PQC

While both protect data against future quantum computers, they approach the problem from completely opposite directions.

Confidence interval (CI)

 The range of values used to estimate a population mean.

Instead of saying:

“The true mean is exactly 50”

statistics says:

“The true mean is probably between 47 and 53.”

That range is the confidence interval.

Central limit theorem (CLT)

It states that the distribution of sample means approximates a Gaussian distribution (normal distribution) as the sample size grows, regardless of the population's original distribution. This is crucial for making inferences about populations based on sample data.

Understanding the CLT can greatly enhance your data analysis skills, providing a solid foundation for hypothesis testing and confidence interval estimation. However, it does have some limitations:

• Sample Size: The CLT requires a sufficiently large sample size (>=30) to be effective. Small samples may not produce accurate results.Independence: The samples must be independent. Dependencies among data points can skew results.
• Identical Distribution: Samples must come from the same distribution. Note: This applies to the classical CLT (Lindeberg-Levy), but newer versions like Lyapunov or Lindeberg-Feller relax this condition.
• Identical Distribution: Samples must come from the same distribution. Note: This applies to the classical CLT (Lindeberg-Levy), but newer versions like Lyapunov or Lindeberg-Feller relax this condition.

The left graph: x-axis represents the actual values (e.g. incomes) of observations in the population. 

y-axis representins probability density or relative probability/frequency.

The right graph is a “distribution of averages,” not a distribution of raw data. X-axis represents the sample mean computed from a sample of size n (out of population size N where n << N). For example:

Step 1: Start with a population

Suppose the population values are:

$$2,\ 4,\ 6,\ 8,\ 10$$

This population may have any shape.

---

Step 2: Take many samples

Take samples of size $n=2$.

Example samples:

• Sample A: $(2,4)$
• Sample B: $(4,10)$
• Sample C: $(6,8)$
• Sample D: $(2,10)$

---

Step 3: Compute a mean for each sample

Each sample produces ONE sample mean:

Sample	Mean
(2,4)	3
(4,10)	7
(6,8)	7
(2,10)	6

So now we have many values of:

$$\bar{x}$$

namely:

$$3,\ 7,\ 7,\ 6,\dots$$

---

Step 4: Plot all those means

The right graph plots the frequencies/probabilities of these sample means.

So the x-axis contains many possible values of:

$$\bar{x}$$

because different samples produce different averages.

คำถามสัมภาษณ์ผู้สมัครเรียนป.เอก

• แนะนำตัวด้านการศึกษาและประสบการณ์ทำงานด้านการศึกษาและประสบการณ์ทำงาน
• ทำไมสนใจเรียนต่อระดับปริญญาเอก จะทำอะไรหลังเรียนจบ
• ถามทำไมสนใจเรียนที่นี่
• ภาระส่วนตัวเยอะไหมเพื่อประเมินความเสี่ยง
• ความเข้าใจในขั้นตอนกระบวนการเรียนต่อปริญญาเอก

Bipartite cache

A bipartite cache is represented as a bipartite graph (aka bigraph), a graph with two disjoint sets of vertices (disjoint set is a set whose no member is shared with other sets) such that every edge connects a vertex in the first set to another vertex in the other set and there are no two vertices in the same set are connected to each other.

Namely proceedings can actually be a journal

Proceedings of the ACM on Measurement and Analysis of Computing Systems (https://www.scimagojr.com/journalsearch.php?q=21101048533&tip=sid&clean=0)

Softmax function

Used in MLP for multiclass classification.

#output nodes = #classes

The output node with highest probability is a returned class. 

https://adeveloperdiary.com/data-science/deep-learning/neural-network-with-softmax-in-python/

Large Multimodal Model (LMM)

A Large Multimodal Model (LMM) represents the next evolution of AI beyond text-only Large Language Models (LLMs). While a traditional LLM is like a brilliant scholar who has only ever read books, an LMM is like that same scholar who can now also see, hear, and create.

At its core, an LMM is a single AI system capable of processing and generating information across multiple "modalities"—such as text, images, audio, and video—all within a unified framework.

CNN vs LSTM

CNNs are capable of extracting spatial correlations among independent variables, while LSTM excels at capturing temporal correlations within input sequences. (https://dl.acm.org/doi/10.1145/3690771.3690774)

The input to CNN must be arranged into 2D array or "feature map" with the shape (Time Steps × Features). Following the input layer is filter (aka kernel, sliding window). With a filter size of 3, this filter looks at 3 time steps and across the features simultaneously. By doing this, the CNN captures how the 12 independent variables interact with each other within a small local window. The input may be 245 time steps and 12 features (i.e., WxL), the first Conv layer may output 245 time steps, 10 features, 32 filters (WxLxDepth), the second Conv layer may output 245 time steps, 8 features, 1 filter. This reduction means the CNN has successfully compressed the 12 feaures into a single optimized "8 feature representation" that carries the most significant spatial information. So the CNN helps "distill" the most important spatial correlations before passing them to the LSTM.

The input to LSTM can have multiple layer as inside an LSTM cell, there are "gates" (Forget, Input, and Output gates). Each gate is essentially a small neural network.

Prompts for generating Notebooklm slide deck

https://github.com/serenakeyitan/awesome-notebookLM-prompts?tab=readme-ov-file#modern-newspaper

Enter prompts into the description box within Notebooklm's Slide deck setting (click right arrow on the button to go to its setting).

Convert image-based slides to editable slides

 https://codia.ai/noteslide

Image-based slides e.g. NotebookLM-generated slides

Hugging face

Hugging Face is a leading open-source AI community and platform often called the "GitHub of Machine Learning." It provides a central hub for researchers and developers to share, discover, and collaborate on AI models, datasets, and applications, with a major focus on Natural Language Processing (NLP) and generative AI.