Sensitivitas terhadap pencilan

Saat menganalisis hubungan lebih dari satu variabel, korelasi adalah langkah awal yang bagus. Namun, bagaimana ketahanan korelasi terhadap himpunan data yang lebih menarik? Seberapa baik korelasi bertahan terhadap pencilan (outlier)?

Dalam latihan ini, Anda akan membuat plot dan menghitung korelasi untuk sebuah himpunan data yang memiliki pencilan, lalu menghapusnya dan melihat apa yang berubah. Pada akhirnya, Anda ingin melihat bagaimana performa korelasi dan menyimpulkan kapan dan di mana Anda sebaiknya menggunakannya.

Contoh himpunan data dari Anscombe's quartet yang terkenal telah diimpor untuk Anda sebagai variabel df, bersama semua paket yang digunakan sebelumnya di bab ini.

Latihan ini adalah bagian dari kursus

Berlatih Pertanyaan Wawancara Statistik dengan Python

Latihan interaktif praktis

Cobalah latihan ini dengan menyelesaikan kode contoh berikut.

# Display the scatter plot of X and Y
plt.scatter(____, ____)
plt.show()

Edit dan Jalankan Kode

Latihan ini adalah bagian dari kursus

Berlatih Pertanyaan Wawancara Statistik dengan Python

SkillTag.level.advancedSkillTag.label

4.7+

Mulai Kursus Gratis

This chapter kicks the course off by reviewing conditional probabilities, Bayes' theorem, and central limit theorem. Along the way, you will learn how to handle questions that work with commonly referenced probability distributions.

Exercise 1: Conditional probabilities Exercise 2: Setting up problems Exercise 3: Bayes' theorem applied Exercise 4: Central limit theorem Exercise 5: Samples from a rolled die Exercise 6: Simulating central limit theorem Exercise 7: Probability distributions Exercise 8: Bernoulli distribution Exercise 9: Binomial distribution Exercise 10: Normal distribution

In this chapter, you will prepare for statistical concepts related to exploratory data analysis. The topics include descriptive statistics, dealing with categorical variables, and relationships between variables. The exercises will prepare you for an analytical assessment or stats-based coding question.

Exercise 1: Statistika deskriptif Exercise 2: Mean atau median Exercise 3: Menghitung simpangan baku secara manual Exercise 4: Data kategorikal Exercise 5: Teknik pengodean Exercise 6: Menjelajahi harga laptop Exercise 7: Dua atau lebih variabel Exercise 8: Jenis-jenis hubungan Exercise 9: Korelasi Pearson Exercise 10: Sensitivitas terhadap pencilan

Latihan Saat Ini

Prepare to dive deeper into crucial concepts regarding experiments and testing by reviewing confidence intervals, hypothesis testing, multiple tests, and the role that power and sample size play. We'll also discuss types of errors, and what they mean in practice.

Exercise 1: Confidence intervals Exercise 2: Confidence interval by hand Exercise 3: Applying confidence intervals Exercise 4: Hypothesis testing Exercise 5: One tailed z-test Exercise 6: Two tailed t-test Exercise 7: Power and sample size Exercise 8: Effect on type II error Exercise 9: Calculating sample size Exercise 10: Visualizing the relationship Exercise 11: Multiple testing Exercise 12: Calculating error rates Exercise 13: Bonferroni correction

Wrapping up, we'll address concepts related closely to regression and classification models. The chapter begins by reviewing fundamental machine learning algorithms and quickly ramps up to model evaluation, dealing with special cases, and the bias-variance tradeoff.

Exercise 1: Regression models Exercise 2: Linear regression Exercise 3: Logistic regression Exercise 4: Evaluating models Exercise 5: Regression evaluation Exercise 6: Classification evaluation Exercise 7: Missing data and outliers Exercise 8: Handling null values Exercise 9: Identifying outliers Exercise 10: Bias-variance tradeoff Exercise 11: Test and training error Exercise 12: Visualizing the tradeoff Exercise 13: Wrapping up