Week 4 - BALT 4363 - Introduction to Linear Algebra for Data Science

There are particular mathematical and scientific principles that underlie the functions of machine learning in AI. I used to believe that only individuals with a quaternary education in mathematics and data science were capable of both understanding and applying these principles in the context of AI. However, after reading about concepts related to linear algebra and data science—including vectors, matrices, and linear transformations—I realized that my previous assumptions were incorrect. Instead of the complex functions I was expecting, simple-to-understand mathematics form the foundations of AI and machine learning. In fact, I recall learning about some of these functions and concepts, such as vectors and matrices, in my junior year of high school.



Despite my familiarity with some of these mathematical ideas, I had never encountered the concept of linear transformations before this reading. Linear transformations can involve processes such as normalization, as demonstrated in the reading’s use of Python. Normalization is a technique that “transforms” data so that it falls within a specific range, typically [0, 1], while preserving the original meaning. This process is especially important when data variables have different units, as it makes it easier to extract critical insights from said data.

Consequently, I began thinking about the potential applications of normalization and Python in my chosen field of international accounting. Differences in systems of measure across countries represent a frequent compatibility issue when trying to extrapolate meaningful insights from international accounting data. The normalization process associated with linear transformations would help to simplify such data. Metric and imperial measures, as well as amounts associated with sales and purchases in different currencies, could be placed on the same scale. As such, comparability and interpretability would increase, and the number of errors would decrease, all of which facilitates the ease, efficiency, and accuracy of deriving critical insights for informed decision-making.

Comments

Post a Comment

Popular posts from this blog

Week 1 - BALT 4363 - Introduction

Image
Hi all!  My name is Anna Lakomiak, and I'm currently a junior at Benedictine University majoring in Accounting. I'm also still deciding between pursuing a minor in Business Analytics or a minor in English and Writing. My interests were not initially inclined toward accounting. Rather, I wanted to pursue a degree in Chemistry as part of a Pre-Pharmacy program. When my senior year of high school rolled around, I realized the career path that I wished to pursue was not for me. It was only when I came to Benedictine University that I joined the Goodwin College of Business and enrolled in an accounting course my freshman year, albeit reluctantly, that I came to discover my love of accounting.   In addition to my studies, I'm involved with activities on campus. Currently, I'm the Secretary of the Accounting Club as well as a Junior Scholar in the Scholars Program. Outside of school, you can find me nestled up with a book or brushing up on my German, Irish, and Swedish languag...
Read more

Week 2 - BALT 4363 - Python Data Manipulation

Image
After delving more into the inner workings of Python, I realized my initial hesitancy regarding this particular programming language was unfounded. I believed that Python would be difficult to work with because of its inherent nature as a programming language. I’ve experimented with different programming languages in the past, all of which were difficult to work with and comprehend. It follows that I would expect Python to be cumbersome as well. Instead, I was quickly surprised to find that it was simple to use and understand. I also found it interesting that many of the different coding functions that Python uses are similar to those of Microsoft Excel. This may be why I found this programming language more intuitive than the ones I’ve experimented with in the past. For example, the “=IF” function in Excel can return a similar output to that of Python’s “if” statements. Furthermore, the “=IF” function in Excel can be combined with the “=OR” function. This combination of Excel func...
Read more

Week 3 - BALT 4363 - Handling and Cleaning Data with Python Libraries

Image
As I continue to learn about the workings of Python, I am again struck by the functional similarities between this computer programming language and Microsoft Excel. These tools can consequently be leveraged, given a particular situation, to maximize both efficiency and meaningful outcomes. When it comes to automation, I personally don’t believe that Python produces quicker results than Excel when data cleaning or basic, functional data tasks are required. There is most likely an equal amount of time spent writing functional code for Python and creating a function or table in Excel in this instance. In the example shown in the picture above, the code necessary to produce the mean global sales for each genre of video game took about as much time, if not more, to write than the time it took me to produce the same results through the PivotTable function in Excel. Despite the advantage Excel offers when simple functions and smaller amounts of data are involved, Python maintains the upper h...
Read more