I recently graduated with a bachelor’s degree in Civil Engineering and was all set to start with a Master’s degree in Transportation Engineering this fall. Unfortunately, my plans got pushed to the winter term because of COVID-19. So as of January this year, I have had no school and no work.

我最近獲得了土木工程學士學位，并準備在今年秋天開始獲得運輸工程碩士學位。 不幸的是，由于COVID-19，我的計劃被推遲到了冬季。 因此，從今年1月起，我沒有上學也沒有工作。

While looking at some of the research going on in my future grad school, I came across Machine Learning and Deep Learning being implemented in a lot of Transportation Engineering related research projects. At the time, I had no clue what ML, DL, or even Data Science as a whole was! So, I started looking into the subject. I talked to a few of my friends who are Computer and Software Engineers and what I understood from them was that Machine Learning is rooted in Statistics, Calculus, and Linear Algebra, all of which are some of my favorite math topics. I remember thinking to myself that I cannot let this opportunity go by, I was in a position where I had all the time in the world and unlimited resources (thanks to the internet!). In addition to that, I was going to enter a field that is being transformed by Data Science very rapidly and I needed to dip my toes into it.

在查看我未來的研究生學校正在進行的一些研究時，我發現機器學習和深度學習已在許多與運輸工程相關的研究項目中實施。 當時，我不知道什么是ML，DL甚至整個數據科學！ 因此，我開始研究該主題。 我與一些計算機和軟件工程師的朋友交談過，我從他們那里了解到，機器學習植根于統計，微積分和線性代數，所有這些都是我最喜歡的數學主題。 我記得自己以為自己不能放過這個機會，當時我處于世界上所有時間無窮無盡的資源(感謝互聯網！)中。 除此之外，我打算快速進入一個由Data Science轉變的領域，我需要全神貫注。

I have always struggled with programming languages, I once started to learn Java but gave it up within 5 hours of starting. It was quite embarrassing as I had high hopes of developing android apps. Now that I think about it, I was just being lazy and impatient. But this time was different, I had to learn Python which is easier to grasp than Java and I found myself truly fascinated by this new field.

我一直在努力學習編程語言，我曾經開始學習Java，但是在開始學習后的5個小時內就放棄了。 由于我對開發android應用程序寄予厚望，這非常令人尷尬。 現在，我開始思考，我只是懶惰而急躁。 但是這次不一樣，我不得不學習比Java更容易掌握的Python，我發現自己對這個新領域非常著迷。

So, let’s jump right in! In this article, I will take you through the journey of how I went from being a complete newbie to a Google Certified TensorFlow Developer in less than 5 months.

所以，讓我們跳進去吧！ 在本文中，我將帶領您完成我如何 在不到5個月的時間內從完全的新手變成了Google認證的TensorFlow開發人員。

1.學習Python (1. Learn Python)

There are a lot of resources available to learn Python from, both free tutorials as well as paid courses which give you a certificate for completion. I personally chose a certificate course as that provided me with a tangible form of credibility and kept me accountable. Coming from a non-coding background this was important to me. Here are some of the resources available;

免費教程和付費課程(可為您提供結業證書)提供了大量學習Python的資源。 我個人選擇了證書課程，因為它為我提供了切實的信譽形式，并讓我負責。 來自非編碼背景，這對我很重要。 以下是一些可用資源；

CERTIFICATE COURSES

證書課程

• This is the course I took to learn Python. I recommend this to anyone who does not have a coding background as this course covers all the fundamentals and is constantly updated. The best part is that you get lifetime access to all the materials and a certificate to prove that you have completed the entire course.
  這是我學習Python的課程。 我推薦給沒有編碼背景的任何人，因為本課程涵蓋了所有基礎知識并且會不斷更新。 最好的部分是您可以終生使用所有材料和證書，以證明您已完成整個課程。

Link: Python Bootcamps: Learn Python Programming and Code Training

鏈接： Python訓練營：學習Python編程和代碼培訓

• This is another wonderful specialization offered by the University of Michigan on coursera.com. It consists of 4 different courses covering a wide range of topics starting from writing your first “Hello World!” code to working with databases. You can either buy it to get the certificate or audit it for free.

  這是密歇根大學在coursera.com上提供的另一個出色的專業。 從撰寫第一個“ Hello World！”開始，它包含4個不同的課程，涵蓋了廣泛的主題。 使用數據庫的代碼。 您可以購買該證書以獲得證書，也可以免費對其進行審核。

Link: Python for Everybody

鏈接： 適用于所有人的Python

List of more courses here

此處有更多課程列表

FREE TUTORIALS

免費教學

• The wonderful people at freeCodeCamp.org regularly post quality coding tutorials on YouTube.

  freeCodeCamp.org的精彩人士定期在YouTube上發布高質量的編碼教程。

Link: Learn Python — Full Course for Beginners [Tutorial]

鏈接： 學習Python —初學者完整課程[教程]

• Another great tutorial video on Python
  另一個關于Python的精彩教程視頻

Link: Python Tutorial — Python for Beginners [Full Course]

鏈接： Python教程—面向初學者的Python [完整課程]

Alright, now you have learnt the basics of Python, great job and congratulations!!

好了，現在您已經學習了Python的基礎知識，出色的工作和恭喜您！！

But, do not expect yourself to be an expert, just because you have a certificate or you sat through a 5 hour long tutorial video. The work is far from done! It is going to be a gradual process and there are some great tools to help you practice and improve your coding skills.

但是，不要僅僅因為擁有證書或坐在5個小時的教程視頻中就期望自己成為專家。 工作還遠遠沒有完成！ 這將是一個循序漸進的過程，其中有些很棒 幫助您練習和提高編碼技能的工具。

I used two websites,

我使用了兩個網站

• Codewars.com is just amazing! they have figured out a way to gamify the process of practicing to code. DO CHECK IT OUT

  Codewars.com真是太神奇了！ 他們找到了一種方法，可以將練習編碼的過程進行游戲化。 檢查一下

Link: Codewars: Achieve mastery through challenge

鏈接： Codewars：通過挑戰實現精通

• Leetcode.com is another great website. They have coding interview style questions in order of increasing difficulty and is another wonderful place to practice and improve your coding skills.

  Leetcode.com是另一個很棒的網站。 他們有按難度遞增的編碼面試風格問題，是練習和提高編碼技能的另一個好地方。

I would recommend to keep practicing and polishing up your Python skills on the side at regular intervals. Remember: spaced repetition works!

我建議您定期定期練習和完善您的Python技能。 請記住：間隔重復有效！

2.學習機器學習理論 (2. Learn Machine Learning Theory)

As I mentioned before, Machine Learning is rooted in Statistics, Calculus and Linear Algebra and hence you do not need to be able to code in order to understand and learn Machine Learning concepts. The Machine Learning course on coursera.com taught by Andrew Ng is an absolute gem in my opinion. The course is old and uses Matlab over Python but the way the concepts are introduced and explained is very relevant, and will prep you well. For a complete newbie like myself, it was extremely frustrating at times but I am grateful now that I completed the course even though I didn’t fully understand some of the topics at the time. I find myself referring to material in the course all the time even while preparing for the TensorFlow Developer Exam.

如前所述，機器學習植根于統計，微積分和線性代數，因此您無需為了理解和學習機器學習概念而進行編碼。 在我看來， Coursera.com上的機器學習課程由Andrew Ng教授，絕對是一門瑰寶。 該課程雖然很老，并且在Python上使用Matlab，但是引入和解釋概念的方式非常相關，可以為您做好準備。 對于像我這樣的完整新手，有時會感到非常沮喪，但是即使我當時還不完全了解某些主題，我也很高興能完成本課程。 即使在準備TensorFlow開發人員考試時，我也總是在參考課程中的內容。

Link: Coursera | Online Courses & Credentials From Top Educators. Join for Free | Coursera

鏈接： Coursera | 來自頂尖教育家的在線課程和資格證書。 免費加入| Coursera

Note: This course can be taken even before learning Python, but I would recommend learning Python first so that you can practice it while learning Machine Learning concepts.

注意：本課程甚至可以在學習Python之前進行，但是我建議您首先學習Python，以便您可以在學習機器學習概念時進行實踐。

3.學習數據科學圖書館 (3. Learn Data Science Libraries)

There are specific libraries within Python that make Data Science related tasks much simpler and efficient. Some of these libraries are Pandas (data manipulation and analysis), Numpy (support for multi-dimensional arrays and matrices), Matplotlib (plotting) and Scikitlearn (creating ML models). There are countless resources available online, here are some of the ones I used -

Python中有特定的庫，這些庫使與數據科學相關的任務更加簡單和高效。 這些庫中有一些是Pandas(數據處理和分析)，Numpy(支持多維數組和矩陣)，Matplotlib(繪圖)和Scikitlearn(創建ML模型)。 網上有無數可用資源，以下是我使用過的一些資源-

• Pandas — I went over a lot of videos, tutorials and even audited a certificate course. This YouTube playlist by codebasics is hands down one the best resources on the internet.
  熊貓-我瀏覽了許多視頻，教程，甚至審核了證書課程。 這個基于codebasics的YouTube播放列表是互聯網上最好的資源之一。

Link: Pandas Tutorial (Data Analysis In Python)

鏈接： Pandas教程(Python中的數據分析)

• Numpy — As usual freeCodeCamp.org for the win!!

  Numpy —像往常一樣免費獲勝！

Link: Python NumPy Tutorial for Beginners

鏈接： Python NumPy初學者教程

• Matplotlib — This particular playlist on YouTube is easy to follow and explains tricky topics very well.
  Matplotlib-YouTube上的此特定播放列表易于遵循，并且很好地解釋了棘手的主題。

Link: Matplotlib Tutorial Series — Graphing in Python

鏈接： Matplotlib教程系列— Python圖形

• Scikitlearn — I took a course offered on udemy.com which covered almost all ML model and explained it’s implementation using real real world data sets. There is also a free 3 hour long tutorial on YouTube which can be found here.

  Scikitlearn-我參加了udemy.com上提供的一門課程，該課程涵蓋了幾乎所有的ML模型，并解釋了它是使用真實世界的數據集實現的。 YouTube上還有一個3小時免費的教程，可以在這里找到。

Link: Machine Learning A-Z (Python & R in Data Science Course)

鏈接： 機器學習AZ(數據科學課程中的Python和R)

4.深度學習理論 (4. Deep Learning Theory)

At this point, it is safe to say that you know most of the things you need to know to become a successful Data Scientist, except Deep Learning. DL is complicated enough that it requires a separate course and even a separate library to implement it. The Machine Learning course by Andrew Ng on Coursera mentions Neural Networks and Deep Learning but there is much more to it than covered in that particular course and so Andrew Ng made another course that just goes deeper into Neural Networks and Deep Learning. Deeplearning.ai created a specialization which consists of 5 courses that cover different topics such as Convolutional Neural Networks, Hyperparameter Tuning, Sequence Models and more. This course focuses on the theory and the under-the-hood working of different Neural Network models. I believe that completing this course is essential even though you don’t absolutely need to in order to successfully build and deploy deep learning models. However, it makes your life much simpler when you have to tune a model or create a model from scratch in TensorFlow.

在這一點上，可以肯定地說，除了深度學習之外，您已經了解成為一名成功的數據科學家所需的大多數知識。 DL非常復雜，因此需要單獨的課程，甚至需要單獨的庫來實現。 吳安德(Andrew Ng)在Coursera上的機器學習課程提到了神經網絡和深度學習，但是它所涉及的內容遠遠超出了該特定課程，因此吳安德(Andrew Ng)開設了另一門課程，它更深入地介紹了神經網絡和深度學習。 Deeplearning.ai創建了一個由5門課程組成的專業課程，涵蓋了不同主題，例如卷積神經網絡，超參數調整，序列模型等。 本課程側重于不同神經網絡模型的理論和后臺工作。 我相信，即使您并非一定要成功構建和部署深度學習模型，也必須完成本課程。 但是，當您必須在TensorFlow中調整模型或從頭開始創建模型時，它會使您的工作變得更加簡單。

Link: Deep Learning Specialization

鏈接： 深度學習專業化

Note: This course can be audited for free.

注意：本課程可以免費審核。

I wrote an article which explains the inner workings of a Deep Neural Network by practical implementation and can be found here

我寫了一篇文章，通過實際實現解釋了深度神經網絡的內部工作原理，可以在這里找到

5. TensorFlow (5. TensorFlow)

TensorFlow is a free and open-source software library used for machine learning applications such as neural networks. Other similar libraries are PyTorch and Theano, but I decided to go forward with TensorFlow as it is supposedly much better for production models and scalability especially since Keras is now completely integrated into TensorFlow. The people at Deeplearning.ai have released another specialization on Coursera which is a continuation of the Deep Learning course mentioned above. It uses the concepts taught in the DL course and implements them using TensorFlow. Another reason I chose to take the TensorFlow in Practice Specialization was that it covered all the pre-requisites required for the TensorFlow Developer Certification Exam by Google.

TensorFlow是一個免費的開源軟件庫，用于機器學習應用程序(例如神經網絡)。 其他類似的庫是PyTorch和Theano，但我決定繼續使用TensorFlow，因為它對于生產模型和可伸縮性據稱要好得多，尤其是因為Keras現在已完全集成到TensorFlow中。 Deeplearning.ai上的人員已經發布了Coursera上的另一個專業化課程，這是上述深度學習課程的延續。 它使用DL課程中講授的概念，并使用TensorFlow實施它們。 我選擇參加TensorFlow實踐專業化的另一個原因是，它涵蓋了Google TensorFlow開發人員認證考試的所有先決條件。

Link: TensorFlow in Practice

鏈接： TensorFlow實踐

DO I NEED TO TAKE THE TensorFlow DEVELOPERS EXAM?

我需要參加TensorFlow開發人員考試嗎？

This particular certificate exam is fairly new, about 5 months old at this point in time, and thus it is not certain right now how valuable it is going to be in the industry with regards to improving job prospects. Coming from a non-coding background this exam acted as a way to validate my skills in Deep Learning.

這項特殊的證書考試還很新，大約五個月之久，因此目前尚不確定它在改善工作前景方面在行業中有多有價值。 來自非編碼背景，此考試是一種驗證我的深度學習技能的方法。

This exam needs a certain level of preparation and coding ability to be able to even attempt it successfully. More than anything else it gave me a set-in-stone goal and acted as motivation to get through all the courses.

該考試需要一定水平的準備和編碼能力，才能成功嘗試。 最重要的是，它給了我一個堅定的目標，并成為我完成所有課程的動力。

Here is a great article talking about the exam in detail.

這是一篇很棒的文章，詳細討論了考試。

These are the steps I followed and I intend to continue learning. There is still a lot more to learn especially since Data Science is such a rapidly developing field.

這些是我遵循的步驟，我打算繼續學習。 特別是因為數據科學是一個發展Swift的領域，所以還有很多東西要學習。