Forem: Hisham Elamir

The Struggle of Producing Machine Learning and AI Research Papers

Hisham Elamir — Mon, 29 Jul 2024 10:43:20 +0000

Introduction

In the rapidly evolving fields of machine learning (ML) and artificial intelligence (AI), the sheer volume of research papers can be overwhelming.

While these papers often showcase groundbreaking advancements, many researchers face significant challenges in replicating the results or building upon the work due to a lack of supporting materials, such as datasets and code.

This article delves into the common struggles -or my struggle at least- faced by researchers in producing high-quality ML and AI papers and explores potential solutions to overcome these obstacles.

Challenges in Producing Research Papers

1. Lack of Supporting Materials

One of the most significant challenges and the main I struggle with when it comes to read or produce a research paper is the absence of supporting materials.
Many research papers provide theoretical insights and results but I don't know why, they don't include the necessary datasets or code to reproduce the experiments.
This lack of transparency hinders other researchers from validating the findings or extending the work. According to a review on the reproducibility of ML research, the major barriers include unpublished data, source code, and the sensitivity of ML training conditions.

2. Complexity of Advanced Techniques

The complexity of modern ML and AI techniques can also be a barrier, hear me on this, the research area became so complex and in absence of material it becomes impossible.
Many state-of-the-art methods require a deep understanding of sophisticated algorithms and access to high-performance computing resources.
For researchers without these resources, replicating or building upon existing work can be daunting. This is particularly true in fields like materials science, where small data sets and specialized algorithms are prevalent.

3. Data Availability and Quality

Data is the cornerstone of ML research, yet obtaining high-quality, relevant datasets can be challenging. While there are many open-source datasets available, not all of them are suitable for specific research needs.
The emphasis on data-centric approaches highlights the importance of quality, diversity, and relevance of data in training ML models. However, finding or curating such datasets can be time-consuming and resource-intensive.

Potential Solutions

1. Utilizing Platforms and Repositories

Platforms like Papers With Code provide a valuable resource by linking research papers with their corresponding code and datasets. This initiative promotes transparency and reproducibility in ML research by making it easier for researchers to access the necessary materials to replicate studies.

2. Leveraging AI Tools for Academic Writing

AI tools can significantly enhance research productivity by assisting with various aspects of academic writing, from idea development and literature review to data management and analysis. These tools can help streamline the research process and improve the quality of the final paper.

3. Fostering a Culture of Open Science

Encouraging a culture of open science where researchers share their data, code, and methodologies can help address the reproducibility crisis. Journals and conferences can play a crucial role by requiring authors to provide supporting materials as part of the publication process. This approach not only aids in reproducibility but also facilitates collaboration and innovation within the research community.

4. Accessing Curated Lists and Resources

Repositories like Awesome Machine Learning on Source Code offer curated lists of research papers, datasets, and software projects related to ML applications. These resources can serve as a starting point for researchers looking to explore specific areas of ML and AI without having to navigate the vast landscape of available literature on their own.

Conclusion

The struggle to produce high-quality ML and AI research papers is a multifaceted challenge, exacerbated by the lack of supporting materials, the complexity of advanced techniques, and issues related to data availability and quality.
However, by leveraging platforms that promote transparency, utilizing AI tools for academic writing, fostering a culture of open science, and accessing curated resources, researchers can overcome these obstacles and contribute valuable insights to the field.

By addressing these challenges head-on, the research community can ensure that ML and AI continue to advance in a way that is both innovative and reproducible, ultimately leading to more robust and impactful scientific discoveries.

Word from me @hishamelamir, I want to write a research and I will go in another way and another approach of doing this, I will share every single step I did to produce, learn and innovate research and making this for you just to learn and go on the same journey. Hopefully you can both learn and teach me if you found new interesting things too.
Thanks for reading 😍.

From Chaos to Clarity: Embracing a 60-Day Commitment on DEV.to

Hisham Elamir — Fri, 26 Jul 2024 17:44:53 +0000

Life is a journey of constant learning and growth, and sometimes, it takes a commitment to truly unlock our potential. My 60-day challenge on DEV.to wasn't just about writing, reading, and learning—it was a transformative experience that taught me discipline, helped me forge meaningful connections, and allowed me to build an audience.

Here's a glimpse into my journey and the lessons I learned along the way.

Beside this article and the upcoming ones, I wrote a free book to describe every single essence and technique I try with good and bad results and feedbacks, here's the link: https://leanpub.com/kanban_your_life

The Beginning: Setting the Stage

The idea of committing to a 60-day challenge might seem daunting at first. For me, it was about more than just honing my writing skills; it was about creating a disciplined routine that would push me out of my comfort zone.
Inspired by the DEV Challenges, which encourage developers to engage with the community through creative and skill-building tasks, I decided to embark on this journey with a clear goal: to write, read, and learn every day for 60 days.

The Discipline of Daily Writing

Writing every day was no small feat. It required me to carve out time from my busy schedule, to sit down and put my thoughts into words. This daily practice not only improved my writing skills but also instilled a sense of discipline that spilled over into other areas of my life. As I wrote about various topics, from technical tutorials to personal reflections, I found myself becoming more organized and focused.

The Joy of Reading

Reading is an integral part of writing. To write well, one must read widely. During my 60-day challenge, I made it a point to read articles from other developers on DEV.to.
This not only provided me with new insights and ideas but also helped me understand different writing styles and techniques. Reading became a source of inspiration and learning, fueling my creativity and broadening my horizons.

Learning and Growing

Learning was at the heart of this challenge. Every day, I sought to learn something new—whether it was a new programming language, a design pattern, or a life lesson shared by a fellow developer.
The process of learning and then writing about it reinforced my understanding and helped me retain the knowledge better. It was a cycle of continuous improvement, where each day built upon the previous one.

Building Connections

One of the most rewarding aspects of this challenge was the connections I made along the way. As I shared my journey and my learnings, I received feedback and encouragement from the DEV.to community.
I met like-minded individuals who were on similar paths, and we supported each other through our respective challenges. These connections turned into friendships, and the sense of community made the journey even more fulfilling.

Gaining an Audience

As I consistently posted my articles, I noticed my audience growing. People started following my work, engaging with my posts, and sharing their thoughts. This not only boosted my confidence but also motivated me to keep going.
Building an audience wasn't just about numbers; it was about creating a space where I could share my knowledge and experiences, and in turn, learn from others.

Reflections on Life

This 60-day commitment was more than just a writing challenge—it was a life lesson. It taught me the value of discipline, the importance of continuous learning, and the power of community.
It showed me that when we push ourselves beyond our limits, we can achieve more than we ever thought possible. Life is about growth, and sometimes, all it takes is a commitment to set the wheels in motion.

Conclusion

Embarking on a 60-day challenge on DEV.to was a transformative experience that went beyond writing, reading, and learning. It was a journey of self-discovery, discipline, and connection.

As I look back, I realize that the true value of this challenge lies not in the articles I wrote or the knowledge I gained, but in the person I became along the way. So, if you're contemplating a similar commitment, I encourage you to take the leap. Embrace the challenge, and let it change your life.

This article is a reflection of my personal journey and the profound impact of a 60-day commitment on DEV.to. It is a testament to the power of discipline, the joy of learning, and the strength of community. Here's to many more challenges and the endless possibilities they bring.

Beside this article and the upcoming ones, I wrote a free book to describe every single essence and technique I try with good and bad results and feedbacks, here's the link: https://leanpub.com/kanban_your_life

From Chaos to Clarity: How Google Keep and Calendar Transformed My Daily Routine

Hisham Elamir — Wed, 24 Jul 2024 07:26:49 +0000

The Messy Beginning

I used to be the epitome of disorganization. My days were a blur of forgotten appointments, misplaced notes, and a constant feeling of being overwhelmed. Post-it notes cluttered my desk, my phone was filled with half-written reminders, and my calendar was a jumble of overlapping events. I knew something had to change, but I didn't know where to start.

Beside this article and the upcoming ones, I wrote a free book to describe every single essence and technique I try with good and bad results and feedbacks, here's the link: https://leanpub.com/kanban_your_life

The Turning Point: Discovering Google Keep and Calendar

One particularly chaotic day, after missing an important deadline and double-booking myself for the third time that week, I decided enough was enough. I needed a system that was always with me, easy to use, and could handle both my tasks and my schedule. That's when I turned to Google Keep and Google Calendar.

Hence, here's the example of my calendar now, I am trying to optimize my daily tasks, for example, I can listen to podcast while doing a routinary task which doesn't require much focus. I calculate the time it takes to drive to work and optimize that.

Step 1: Centralizing My Notes with Google Keep

I started by downloading the Google Keep app on my phone and opening it on my computer. Here's what I did:

Migrated All Notes: I spent an evening transferring all my scattered notes into Google Keep. It was tedious, but seeing everything in one place was instantly relieving.
Color Coding: I used different colors for different types of notes. Work-related notes were blue, personal tasks were green, and ideas were yellow.
Labels: I created labels like "Work," "Personal," "Ideas," and "Urgent" to further categorize my notes.
Archiving: I learned to archive notes that weren't immediately relevant, keeping my main screen clutter-free while still having access to all my information.

Step 2: Mastering Task Management

Next, I focused on turning my notes into actionable tasks:

Creating Checklists: For any note that involved multiple steps, I converted it into a checklist in Google Keep.
Setting Reminders: I added date and time reminders to my most important tasks, ensuring I wouldn't forget them.
Integrating with Calendar: I discovered that tasks with due dates automatically appeared in my Google Calendar, creating a seamless system.

Step 3: Optimizing My Calendar

With my notes and tasks organized, I turned my attention to my Google Calendar:

Time Blocking: I started allocating specific time blocks for different types of tasks, using the colors that matched my Google Keep categories.
Task Integration: I enabled the Tasks view in Google Calendar, allowing me to see my to-do list alongside my appointments.
Recurring Events: I set up recurring events for regular commitments, reducing the need to manually input repeating tasks.

The Results: A More Structured and Productive Day

After a few weeks of consistently using this system, the change was dramatic:

Improved Time Management: With all my tasks and appointments in one place, I could see at a glance what my day looked like and plan accordingly.
Reduced Stress: No more forgotten tasks or double-booked appointments. Everything was in the system, and I could trust it.
Increased Productivity: By focusing on one task at a time and having a clear view of my priorities, I found myself completing more work with less stress.
Better Work-Life Balance: Separating work and personal tasks visually helped me maintain boundaries and ensure I was allocating time for both.

Conclusion: Small Steps, Big Impact

My journey from chaos to clarity didn't happen overnight. It started with the small step of centralizing my notes in Google Keep, then gradually expanded to full calendar integration. The key was consistency and a willingness to adjust the system as I learned what worked best for me.

If you're feeling overwhelmed by disorganization, I encourage you to start small. Pick one tool, like Google Keep, and commit to using it for all your notes for a week. Then, gradually expand your system. Remember, the goal isn't perfection, but progress. With each small step, you'll find yourself moving from chaos to clarity, just as I did.

Beside this article and the upcoming ones, I wrote a free book to describe every single essence and technique I try with good and bad results and feedbacks, here's the link: https://leanpub.com/kanban_your_life

🐍 X 10 Advanced Python Hacks 4 You

Hisham Elamir — Tue, 23 Jul 2024 11:37:28 +0000

Python is a versatile and powerful language, and mastering its advanced features can significantly enhance your coding efficiency and readability. Here are some advanced Python tips to help you write better, cleaner, and more efficient code.

I wrote 2 small books to read in weekend that covers python, here's the links: (1) https://leanpub.com/learnpython_inweekend_pt1 & (2) https://leanpub.com/learnpython_inweekend_pt2

1. Use List Comprehensions for Concise Code

List comprehensions offer a concise way to create lists. They can often replace traditional for-loops and conditional statements, resulting in cleaner and more readable code.

# Traditional approach
numbers = [1, 2, 3, 4, 5]
squared_numbers = []
for num in numbers:
    squared_numbers.append(num ** 2)

# Using list comprehension
squared_numbers = [num ** 2 for num in numbers]

2. Leverage Generator Expressions for Memory Efficiency

Generator expressions allow you to create iterators in a concise manner without storing the entire sequence in memory, making them more memory-efficient.

# List comprehension (creates a list)
squared_numbers = [num ** 2 for num in numbers]

# Generator expression (creates an iterator)
squared_numbers = (num ** 2 for num in numbers)

3. Utilize `enumerate()` for Index Tracking

When iterating over an iterable and needing to track the index of each element, the enumerate() function is invaluable.

fruits = ['apple', 'banana', 'cherry']
for index, fruit in enumerate(fruits):
    print(f"Index: {index}, Fruit: {fruit}")

4. Simplify String Concatenation with `join()`

Using the join() method to concatenate strings is more efficient than using the + operator, especially for large strings.

fruits = ['apple', 'banana', 'cherry']
fruit_string = ', '.join(fruits)
print(fruit_string)  # Output: apple, banana, cherry

5. Use `slots` to Reduce Memory Usage

By default, Python stores instance attributes in a dictionary, which can consume significant memory. Using __slots__ can reduce memory usage by allocating memory for a fixed set of instance variables.

class Point:
    __slots__ = ['x', 'y']
    def __init__(self, x, y):
        self.x = x
        self.y = y

6. Employ `contextlib.suppress` to Ignore Exceptions

The contextlib.suppress context manager allows you to ignore specific exceptions, simplifying your code by avoiding unnecessary try-except blocks.

from contextlib import suppress

with suppress(FileNotFoundError):
    with open('file.txt', 'r') as file:
        contents = file.read()

7. Utilize the `itertools` Module

The itertools module offers a collection of efficient functions for working with iterators. Functions like product, permutations, and combinations can simplify complex operations.

import itertools

# Calculate all products of an input
print(list(itertools.product('abc', repeat=2)))

# Calculate all permutations
print(list(itertools.permutations('abc')))

8. Use `functools.lru_cache` for Caching

The functools.lru_cache decorator can cache the results of expensive function calls, improving performance.

from functools import lru_cache

@lru_cache(maxsize=32)
def fibonacci(n):
    if n < 2:
        return n
    return fibonacci(n-1) + fibonacci(n-2)

9. Master Decorators for Cleaner Code

Decorators are a powerful tool for modifying the behavior of functions or classes. They can be used for logging, access control, and more.

def my_decorator(func):
    def wrapper():
        print("Something is happening before the function is called.")
        func()
        print("Something is happening after the function is called.")
    return wrapper

@my_decorator
def say_hello():
    print("Hello!")

say_hello()

10. Use the For-Else Trick

The for-else construct in Python allows you to execute an else block after a for loop completes normally (i.e., without encountering a break statement). This can be particularly useful in search operations.

for n in range(2, 10):
    for x in range(2, n):
        if n % x == 0:
            print(f"{n} equals {x} * {n//x}")
            break
    else:
        # Loop fell through without finding a factor
        print(f"{n} is a prime number")

Conclusion

By incorporating these advanced Python tips into your development workflow, you can write more efficient, readable, and maintainable code.

Whether you're optimizing memory usage with __slots__, simplifying string operations with join(), or leveraging the power of the itertools module, these techniques can significantly enhance your Python programming skills.

Keep exploring and practicing these concepts to stay ahead in your Python journey.

From Chaos to Clarity: How Kanban Transformed My Daily Routine

Hisham Elamir — Tue, 23 Jul 2024 11:20:26 +0000

I am a Machine Learning Engineer and my life is going well, but lately I am feeling that there's a big time issue on my day and I have a lot in my mind that I end my day without finishing it, so I tried a work management technique which is agile to transform my life.

Introduction

I used to live in a constant state of chaos. My days were a blur of unfinished tasks, missed deadlines, and a growing sense of frustration.

I felt overwhelmed and disorganized, unable to keep up with the demands of both my personal and professional life. Then, I discovered Kanban—a simple yet powerful system that transformed my daily routine from chaos to clarity.

Beside this article and the upcoming ones, I wrote a free book to describe every single essence and technique I try with good and bad results and feedbacks, here's the link: https://leanpub.com/kanban_your_life

The Turning Point: Discovering Kanban

Kanban, a Japanese term meaning "signboard" or "billboard", was originally developed by Toyota in the 1940s to optimize production processes.

I watched historical movie about Toyota, and actually I learnt a lot of things that can help to make work/life more productive.

The core idea is to visualize work and manage it efficiently by moving tasks through different stages of completion. Inspired by its success in manufacturing and software development, I decided to apply Kanban principles to my own life.

Setting Up My Personal Kanban Board

Creating my personal Kanban board was the first step towards regaining control over my daily routine. Here's how I did it:

Choosing the Medium: I opted for a physical whiteboard, but you can also use digital tools like Trello or Asana.
Defining Columns: I divided the board into three columns: "To Do," "In Progress," and "Done." This simple structure allowed me to categorize my tasks and visualize my workflow.
Adding Tasks: I wrote down all my tasks on sticky notes and placed them in the "To Do" column. This included everything from work projects to household chores.
Limiting Work-in-Progress (WIP): To avoid feeling overwhelmed, I set a limit of three tasks in the "In Progress" column at any given time. This helped me focus on completing tasks before starting new ones.

The Transformation: From Overwhelm to Organized

The impact of implementing Kanban was immediate and profound. Here are some of the key benefits I experienced:

Clarity and Focus: Visualizing my tasks on the Kanban board reduced mental clutter and provided a clear overview of my workload. I could see at a glance what needed to be done, what was in progress, and what was completed.
Improved Productivity: By limiting my work-in-progress, I was able to concentrate on finishing tasks more efficiently. This reduced the stress of multitasking and increased the quality of my work.
Motivation and Satisfaction: Moving tasks from "To Do" to "Done" gave me a tangible sense of progress and accomplishment. This visual feedback loop kept me motivated and encouraged me to keep going.
Better Time Management: The structure of the Kanban board helped me prioritize tasks and manage my time more effectively. I developed a habit of sticking to deadlines and meeting my goals.
Work-Life Balance: Using separate boards for work and personal tasks allowed me to manage both areas of my life without feeling overwhelmed. I was able to allocate time for family, hobbies, and self-care.

Practical Tips for Using Kanban in Daily Life

If you're considering implementing Kanban in your own life, here are some practical tips to get started:

Start Simple: Begin with a basic board and a few columns. You can always add more complexity as you become comfortable with the system.
Be Consistent: Use the Kanban board daily and update it regularly. Consistency is key to reaping the benefits of the system.
Set Realistic WIP Limits: Experiment with different WIP limits to find what works best for you. The goal is to avoid overwhelm and maintain focus.
Break Down Tasks: Divide larger tasks into smaller, manageable steps. This makes it easier to track progress and stay motivated.
Review and Adapt: Regularly review your board and make adjustments as needed. Kanban is meant to be flexible and adaptable to your changing needs.
Share Results: One of the things that made me do commit on this hard lifestyle is sharing, I shared with smaller circle of friends and discussed with them the daily blockers.

Conclusion

Kanban transformed my daily routine from a state of chaos to one of clarity and control. By visualizing my tasks, limiting work-in-progress, and managing my workflow, I was able to boost my productivity, reduce stress, and achieve a better work-life balance.

If you're feeling overwhelmed and disorganized, I encourage you to give Kanban a try and write a small comment to discuss you ideas of enhancement. It might just be the solution you need to bring order to your life and help you reach your full potential.

Beside this article and the upcoming ones, I wrote a free book to describe every single essence and technique I try with good and bad results and feedbacks, here's the link: https://leanpub.com/kanban_your_life

🐍 Python Pro Tips That I Wish I Use More...

Hisham Elamir — Mon, 22 Jul 2024 06:11:04 +0000

As a Python enthusiast, I'm always on the lookout for ways to write cleaner, more efficient code. Here are some of my favorite Python tricks that can help you become a more effective programmer:

List Comprehensions: Create lists with concise, readable one-liners.

even_numbers = [x for x in range(10) if x % 2 == 0 and x != 0]

F-strings: Format strings easily and readably (Python 3.6+).

name, age = "John", 30
print(f"I'm {name} and I'm {age} years old.")

Enumerate: Get both index and value when iterating.

for index, value in enumerate(["a", "b", "c"]):
    print(f"Index: {index}, Value: {value}")

Unpacking: Assign multiple variables in one line.

a, b, c = 1, 2, 3

Slicing: Manipulate sequences efficiently.

my_list = [0, 1, 2, 3, 4, 5]
reversed_list = my_list[::-1]

Use Counter: use collections.Counter for counting elements.

from collections import Counter
my_list = [1, 1, 2, 3, 3, 3, 4]
print(Counter(my_list))

Flattening: Flatten nested lists with itertools.chain.

import itertools
nested_list = [[1, 2], [3, 4], [5, 6]]
flat_list = list(itertools.chain.from_iterable(nested_list))

Iterate Over Many: Use zip() to iterate over multiple lists simultaneously.

names = ["Alice", "Bob", "Charlie"]
ages = [25, 30, 35]
for name, age in zip(names, ages):
    print(f"{name} is {age} years old")

These tricks can help you write more Pythonic code and boost your productivity. What are your favorite Python tips? Share in the comments below!

Thanks for reading.

First Post Lasts

Hisham Elamir — Tue, 14 May 2024 09:17:32 +0000

Hi everyone, Hope that you are having a good week of productivity and learning.
I am writing this post, and I know it's hard to write a good post or article, but also, I am sure that learn new things is much harder. While the joy of learning things can't be overstated.

I am writing to tell you that I am new to writing on DEV community. I love sharing knowledge, no I am wrong. I love talking to people and discuss concepts and philosophy.
And that's why I am here. To share, to discuss and to talk.

I am a Data Scientist and Machine Learning Engineer. But also, I am a bit nerd when it comes to weird things.

I will share few articles this week, to use my motivation and get some momentum. So, I am asking you to give a push, give me momentum to create a small community.

Thanks a lot for reading this post, while it won't give you any useful knowledge.
Hisham.

Forem: Hisham Elamir

The Struggle of Producing Machine Learning and AI Research Papers

Introduction

Challenges in Producing Research Papers

1. Lack of Supporting Materials

2. Complexity of Advanced Techniques

3. Data Availability and Quality

Potential Solutions

1. Utilizing Platforms and Repositories

2. Leveraging AI Tools for Academic Writing

3. Fostering a Culture of Open Science

4. Accessing Curated Lists and Resources

Conclusion

From Chaos to Clarity: Embracing a 60-Day Commitment on DEV.to

The Beginning: Setting the Stage

The Discipline of Daily Writing

The Joy of Reading

Learning and Growing

Building Connections

Gaining an Audience

Reflections on Life

Conclusion

From Chaos to Clarity: How Google Keep and Calendar Transformed My Daily Routine

The Messy Beginning

The Turning Point: Discovering Google Keep and Calendar

Step 1: Centralizing My Notes with Google Keep

Step 2: Mastering Task Management

Step 3: Optimizing My Calendar

The Results: A More Structured and Productive Day

Conclusion: Small Steps, Big Impact

🐍 X 10 Advanced Python Hacks 4 You

1. Use List Comprehensions for Concise Code

2. Leverage Generator Expressions for Memory Efficiency

3. Utilize enumerate() for Index Tracking

4. Simplify String Concatenation with join()

5. Use __slots__ to Reduce Memory Usage

6. Employ contextlib.suppress to Ignore Exceptions

7. Utilize the itertools Module

8. Use functools.lru_cache for Caching

9. Master Decorators for Cleaner Code

10. Use the For-Else Trick

Conclusion

From Chaos to Clarity: How Kanban Transformed My Daily Routine

Introduction

The Turning Point: Discovering Kanban

Setting Up My Personal Kanban Board

The Transformation: From Overwhelm to Organized

Practical Tips for Using Kanban in Daily Life

Conclusion

🐍 Python Pro Tips That I Wish I Use More...

First Post Lasts

3. Utilize `enumerate()` for Index Tracking

4. Simplify String Concatenation with `join()`

5. Use `slots` to Reduce Memory Usage

6. Employ `contextlib.suppress` to Ignore Exceptions

7. Utilize the `itertools` Module

8. Use `functools.lru_cache` for Caching