• Create entries regularly to track your research journey. These should be both text and video (at least one video a week).

• Include your current research stage (e.g., Topic Selection, Literature Review, Method Development)

• Share any questions or concerns you have about your research

• Attach relevant documents or resources that support your work

• Reflect on your progress and learning experiences

This timeline format will help you visualize your research development throughout the year.

Knowing what type of language and jargon to use in my paper will make it more interesting and engaging for my audience. Thinking in terms of values and interests also leads to a more direct and focused paper, as it pertains to my topic. I feel that I'll convey the most important information through thinking in this way.

Who is your intended audience for your project, and what might they care about most?

My intended audience contains experts in the field of machine learning and civil engineering. Specifically, I'm writing this for professionals in pavement auditing who might be able to utilize any information about machine learning as it pertains to efficiencies in determining PCI.

This is one of the potential sources for my project. The main benefit it adds is as a comprehensive research review and evaluation for various ML models. Not to mention, it is directly geared at PCI prediction.

I had previous knowledge of this source as the author was a graduate assistant to Dr. Stephanie Ivey, whom I shadowed under last during during the WBL sessions.

I understand this source quite well and have read it in its entirety. It has the potential to serve as my most important source, as it directly inspired my question.

The care-factor with my project is actually very easy to align with what people would care about. Simply put, i'm contributing to a field which saves money and labor.

Paul and I recently had a debate about who the best rapper of all time was. Here was his argument:

Drake is the best rapper of all time through the combination of talent, sound quality, and general influence in the rapper ethos. He has one of the most successful careers and generally sounds good.

Here was my argument:

No. Drake sounds like a Starbucks barista. I believe kanye is the best rapper of all time due to his commercial success, influence, and voice/lyricism. Not only is he insanely popular, but he produces and writes masterpieces---or at least he used to.

Reflection

• I have put in minimal thought in term of summarizing my sources. Truthfully, I've only summarized them because of the annotated bibliography. I feel that I learn better through in-moment note taking and review, rather than through summarization.

• Because my project is a meta-analysis of sources, I have sparcely thought in terms of the "believing game." I think, and frankly am now realizing as I type this sentence, that summarizing different viewpoints and outcomes will be very beneficial for my research.

I think the main method of research I will employ is through a meta-analysis. I will not produce any in-the-filed type research like a survey or study, as both the research type and materials that I have access to disqualify these outlets. Rather, I will find ongoing and past studies which relate to my research question and report on their findings overall in a manner which synthesizes the data as a whole sum.

Challenges:

While finding studies related to machine learning and civil infrastructure will not be difficult, I think synthesizing the data will be very challenging--specifically through determining and defining the ratio between computing power and accuracy.

The main difference, in terms of how I differentiate the two terms, is that a lit review finds the gap in research, and a meta-analysis answers that gap. They are very interconnected, being essentially two sides of the same coin in terms of the problem solving method.

Moving forward, I will be cognizant of that difference. The first step, I feel, is to research established meta-analyses on Machine Learning, and noting how they conduct research and gather conclusions from that research.

My question has changed more than I anticipated. I've found myself focusing more on a comparison of the performance metrics of AM+CNN vs CNN more than the pavement field or what I was originally planning--A feasibility analysis.

I think that this will actually help me in the loing run. I have hard data that I can run with, adn precision percentages that I can calculated very easily from numerous studies. The outlook of this project is good, in other words.

I think the biggest revisions I need to implement are with the explanation side of this project. I need to ensure that my audience understands what machine learning and PCI is fully before delving into backpropogations and skip connections.