Statistical Machine Learning is used to discover patterns and relationships in large data sets. Topics will include: data exploration, classification and regression tress, random forests, clustering and association rules. Building predictive models focusing on model selection, model comparison and performance evaluation. Emphasis will be on concepts, methods and data analysis; and students are expected to complete a significant class project, individual or team based, using real-world data.
4
UnitsOptional
Grading1, 2
PasstimeNone
Level LimitLetters and science
CollegeProf Baracaldo is not the best at explaining 131 material; she goes rly deep into all of statistical ML proofs and is often times confusing. She's nice though and gives good feedback on project if you approach her after class. HW and quizzes are not bad. TA is rly helpful. Overall not a hard class but ML concepts are hard to understand in general.
Super nice professor, class was very fairly graded on easy homework and final project, clear grading criteria. Available after class and at office hours to answer any and all questions, will help you directly with any specific problem. Quizzes were very simple and open book/internet, just a basic check to make sure you're paying attention.
She knows a lot but can not express it in understandable way, the lectures were not organized, slides are incomplete because she writes additional notes during lecture. No clue what to focus on before final and midterms and didn't even follow her own syllabus. Take her course if you want to lower you GPA and waste time.
I am the PSTAT 115 student.the best professor I have ever met at UCSB. Well-prepared and organized lecture The exams are not easy, but if you follow her step, you will do well on the exams. she should not got low grade, she is so patient when answer my question.I got 100% on canvas, it gives me confidence and I really love it and the professor
4 HW's with 1.5 weeks to finish each, 3 online quizzes bi-weekly. Lectures were so confusing but necessary to succeed so attend all. Midterm and final both took questions from the practice exam and had 50% R output interpretation 30% Derivation/Proof 20% MC/TF and extra credit. Attend sections right before the exam, TA's went over helpful topics.
If you want to be concerned about whether you will graduate on time due to a single question on a single exam, this is the class to take. Prof never responds to emails/Nectir (Slack).