Engineering Maths Graduate
Send a job offer directly to this candidate
I am highly interested in data science and have been since the first few data science projects completed as part of my degree. There is something really satisfying about starting with a monolithic opaque block of data and somehow pulling repeatable, beneficial, actionable conclusions. To me, it feels like successfully drawing blood from a stone. Often requiring a satisfying puzzle that tiptoes the line between computer science and maths, both of which are some of my core interests.
Frequently completing data science modules made up a large part of my course. With 3-4 projects each year. Each required spending 6-8 weeks modelling and then writing a report. The projects were on complex problems, resembling Kaggle challenges, in 3rd year these were real-life problems given to models by a sponsoring business. Examples of these projects include:
Creating a model to predict change to vote distribution of a party dropping out of an election using British election study data. Achieving a model that was 95% accurate when tested on historical survey data
Achieving an F1 score of 0.85, when building a classifier for a specific species of whale from audio data of whale calls. This was completed using a compound model combining a DNN on a large extracted then filtered feature set, combined with a CNN on a spectrogram.
3 years of frequently completing data science projects has given me good experience with tools, such as clustering algorithms, dimensionality reduction, and machine learning techniques. I am especially well-versed in feature extraction-based modelling. I am also quite up to date on new tools being developed. Through these projects, I've also worked in lots of classes of data science problems, and data, be it Time series, visual, audio, or features.
Frequently completing data science modules made up a large part of my course. With 3-4 projects each year. Each required spending 6-8 weeks modelling and then writing a report. The projects were on complex problems, resembling Kaggle challenges, in 3rd year these were real-life problems given to models by a sponsoring business. Examples of these projects include:
Creating a model to predict change to vote distribution of a party dropping out of an election using British election study data. Achieving a model that was 95% accurate when tested on historical survey data
Achieving an F1 score of 0.85, when building a classifier for a specific species of whale from audio data of whale calls. This was completed using a compound model combining a DNN on a large extracted then filtered feature set, combined with a CNN on a spectrogram.
3 years of frequently completing data science projects has given me good experience with tools, such as clustering algorithms, dimensionality reduction, and machine learning techniques. I am especially well-versed in feature extraction-based modelling. I am also quite up to date on new tools being developed. Through these projects, I've also worked in lots of classes of data science problems, and data, be it Time series, visual, audio, or features.
