thesis

1. format
2. accurate bookkeeping
3. thesis as discussion
- 3.1. implications
  - 3.1.1. thesis could be from anything
  - 3.1.2. thesis is discussion with thesis work
Backlinks
- old index
  - (Academically speaking)

Thesis is discussion with:

an conventional thesis 1
2

Below is some elaborated explanations:

1. format

typically, Thesis consist of the following parts:

abstract
introduction
body
conclusion

Following is a detailed description of each section

1.1. abstract

abstract is a self-contained block of text that describes this thesis.

The goal of abstract is to describe what is done in this thesis so the reader could decide whether to read it without spending time reading the whole text.

1.2. introduction

introduction is the warm-up of discussion. It typically describes the context where the discussion–the main event of the thesis–took place.

Typically, there will be a literature review(sometimes even after the thesis is done, so hypocritical). Notation table and definitions are sometimes present. Hypotheses is sometimes present.

literature review: literature refers to other people’s thesis in the past. The purpose is to familiarize the reader of the historical context of the discussion: what progress/discussions have been made, on what ground this discussion is made. It should also list out tool concepts the thesis exploited so reader could check out
Notation table: a simple table noting all notations used in the thesis to avoid confusion/make it easier for reader to keep track of the notations.(they are going to make one themselves anyway)
Definitions: similar to Notation table, Clarify every term/jargon used in the discussion. Simple reference or dependency check if the reader is not familiar with some big terms used.
Hypotheses: Sometimes the purpose of the thesis is to prove some hypotheses, then it’s appropriate to list them out first.

Sometimes important terms would appear here as section.

1.3. body

The main event in the thesis: to discuss a subject matter.

It take many forms, and the structure differs by how you discuss it.

1.4. conclusion

conclusion is where you put forward your conclusion.

Sometimes people would summarize the whole thesis here too, just like 1.1, only that you don’t have to be self-contained, and could write more.

2. accurate bookkeeping

Accurate bookkeeping refers to the convention that:

thesis should show credit for ideas exploited that were obtained from other thesis, and
thesis should keep these records accurate and provide information to retrive them.

They are implemented by the various reference styles

3. thesis as discussion

The most important insight about thesis is that Thesis is discussion.

To be more clear, this implies that thesis could be about anything, as it is with discussion. Any of your idea, your talk with a friend could be the 1.3 of a thesis, with proper working on them.

3.1. implications

This insight have some important implications

3.1.1. thesis could be from anything

I thought of neural network could be a way of constructing functions and it work by combining incorprating more influence and gradient descent to make the influence affect the target function when trying to explain it to my dad –>
I gether literatures about this idea –>
I push it a little more, making more examples, instances(other function constructing mechnism? other neural network component as functions?) –>
with bookkeeping, I pull together the notes to form the literature review and references, and I note where I used other thesis’ results –>
it’s a thesis now.

3.1.2. thesis is discussion with thesis work

To be exact, thesis work is doing literature review, reference bookkeeping and designing and doing experiments(if present). Discussion is the easy part, and actually the daily part for scholars–they could have 2-3 new ideas every day, when reading, talking or taking shower. They might have finished the discussion in the shower(of course, for complicated ones with pen and paper).

Another possible thesis work is exploratory experiments, the purpose of which is not to test hypotheses but to inspire some. In exploratory experiments, some procedures are taken to produce some data/observations, upon which the scholar is likely to spot new patterns and found new idea/concepts. An example is:

I realize “neural network could be a way to construct functions”
I write down in this perspective, how function is constructed with numorous neural network components
in the process, I defined some architecture in this manner
in the process, reflecting on the precedure I illustrated, I find some new ideas that complements the original idea
description make discussion possible. By spanning more description after each discussion, I make progress to more complex/general/useful/subtle subject matters.