Why Study with PlanetCert?
The Latest Questions
Practice questions and exam topics aligned with the current exam objectives.
Detailed Explanations
Go beyond the answer. Master the material with comprehensive learning and professional explanations for every concept.
AI-Powered Insights
Personalized preparation guidance that adapts to your performance and identifies weak spots automatically.
Exam Information
Official specifications published by CIW
Exam Format
Registration
Validity
Skills this practice test covers
- Describe a dataset with descriptive statistics and frequency distributions.
- Choose features and models, and tell overfitting from underfitting.
- Rate a classifier with accuracy, the ROC curve, AUC and F1.
- Fetch data with SQL, then clean out missing values and outliers.
- Analyze a dataset in Python with Pandas and seaborn.
- Run an A/B test and explain its result to non-statisticians.
Explore related topics
1D0-184 exam content
4 domains · 14 topics · 150 questions
Data Science Overview31 questions
This domain defines the field. Candidates explain machine learning, how data science differs from AI and where businesses use it, then weigh the legal, ethical and privacy questions that data work raises, such as biased predictions. It also covers how a data scientist keeps up with new research.
Skills tested
- Distinguish data science from AI and list its business applications.
- Explain what a correlation coefficient is and how it is calculated.
- Explain the harm that biased predictions by data models can cause.
- Apply ethical guidelines and data protection methods to a data science scenario.
- Follow emerging analytic trends through scientific articles and conference papers.
Questions by topic
- 1.1: Fundamentals14
- 1.2: Legal, Ethics and Privacy Considerations15
- 1.3: Career2
Analysis55 questions
This domain holds the most objectives. Candidates explore data with descriptive statistics and clustering, choose features and models, measure model performance and check model assumptions.
Skills tested
- Describe a distribution by its mean, variance, skewness and kurtosis.
- Use clustering and other data mining techniques on a dataset.
- Reduce dimensions with PCA and explain the curse of dimensionality.
- Tell overfitting from underfitting and explain the bias-variance trade off.
- Judge a classifier with accuracy, the ROC curve, AUC and F1.
Questions by topic
- 2.1: Exploratory Data Analysis10
- 2.2: Modeling and Visualization Techniques26
- 2.3: Statistics19
Managing Data28 questions
This domain is about getting data ready to use. Candidates work with data warehouses, distributed systems such as Hadoop, SQL queries and indexes, then clean the data by handling missing values, outliers and categorical variables.
Skills tested
- Write SQL queries to fetch data, and use indexes to make them efficient.
- Explain how Hadoop and MapReduce analyze big datasets.
- List the stages of the data cycle and the methods for acquiring data.
- Handle missing values, outliers and categorical variables before modeling.
- Normalize data as part of preprocessing.
Questions by topic
- 3.1: General Data Management13
- 3.3: Data Preparation11
- 3.2: Querying Databases4
Professional Skills36 questions
This domain covers the work around the analysis. Candidates program in Python and R, run surveys and A/B tests, deploy and update models, and present results to people who are not statisticians.
Skills tested
- Represent data with Pandas and plot it with matplotlib or seaborn.
- Explain computational complexity in algorithm design.
- Design an A/B test and split data into training and testing sets.
- Choose a method for deploying a machine learning model, and update it with fresh data.
- Report results in charts and plain language for non-statisticians.
Questions by topic
- 4.2: Conduct Research5
- 4.1: Programming15
- 4.4: Communicating Results6
- 4.5: Deploy Models8
- 4.6: Problem Identification2
Requirements
- No formal prerequisites: anyone can book the exam.
- Recommended: working knowledge of statistics and probability, and of Python or R.
- Recommended: an understanding of databases and SQL.
- Recommended: the CIW Artificial Intelligence Associate certification first.
About the 1D0-184 exam
1D0-184 is a CIW exam for candidates who already have a grounding in statistics and code. It checks that they can explore data, build models and explain the results. CIW recommends basic statistics and probability, Python or R, SQL, and the CIW Artificial Intelligence Associate certification first, though none of them is required. Its four domains run from data science basics and ethics to statistics, data management and professional skills such as programming and deploying models. Passing earns the CIW AI Data Science Specialist certification.
Analysis has the most questions in the PlanetCert practice test, led by modeling and visualization techniques. The explanations show the reasoning behind each answer, such as why a model that scores well on training data but poorly on test data is overfitting. Study Mode suits statistics topics such as the bias-variance trade off, where each explanation needs a careful read. Brainy AI Assistant can explain a metric such as AUC with a small example, or show how a Pandas command answers a data question.
How to earn the certification
- Earns
- CIW AI Data Science SpecialistSpecialist
The certification is valid for three years and renews when the holder passes the exam again.
- Before
- 1D0-181 CIW Artificial Intelligence Associate
Recommended by CIW before this exam, though not required.
- 1D0-181 CIW Artificial Intelligence Associate
How to prepare for 1D0-184
- 1
Set up Python and R side by side
Install Python with Pandas, matplotlib and seaborn, and install R. The Programming topic expects work in both languages.
- 2
Explore one public dataset fully
Load a dataset, describe each variable, plot its distributions and look for outliers and missing values. This one exercise touches much of Analysis and Managing Data.
- 3
Build and judge two models
Fit a classification model and a regression model. Judge the classifier with accuracy, the ROC curve, AUC and F1, and note where each model overfits or underfits.
- 4
Query a real database with SQL
Write SQL queries that fetch data from several tables, then see how an index changes their efficiency.
- 5
Review ethics, then go timed
Read the legal, ethics and privacy objectives closely, because they need judgment rather than calculation. Then take the 1D0-184 practice test in Timed Mode and ask Brainy AI Assistant about any statistics question that went wrong.
Start a timed 1D0-184run
CIW offers an optional CIW AI Data Science Specialist course for this exam.
Who 1D0-184 is for
- Data analysts who want to move into modeling and machine learning work.
- People who hold CIW Artificial Intelligence Associate and want a hands-on data science credential.
- Programmers who know Python or R and want to apply it to statistics and data.
Typical job titles: Data scientist · Data analyst · Junior data scientist · Business intelligence analyst · Machine learning analyst
1D0-184 glossary
18 terms from the exam objectives and the vendor's official material, each defined in a sentence.
- A/B test
- An experiment that compares two versions on randomly split groups, to measure the effect of one change.
- AUC
- The area under the ROC curve: one number that sums up how well a classifier separates the classes.
- Bias-variance trade off
- The balance between a model too simple to catch the pattern (high bias) and one too sensitive to its training data (high variance).
- Causal inference
- Working out whether one thing actually causes another, not only whether the two move together.
- Correlation coefficient
- A number from -1 to +1 that shows how strongly two variables move together in a straight-line relationship.
- Curse of dimensionality
- The problems that appear as a dataset gains features: the data becomes sparse, and models need many more examples to learn well.
- Exploratory data analysis
- An early study of a dataset with summary statistics and plots, which shows its shape, gaps and odd values before any model is built.
- Feature selection
- Choosing the input variables that help a model most, and dropping the ones that add noise.
- Generalized linear models
- An extension of linear regression that models outcomes such as counts or yes-or-no answers by passing the linear prediction through a link function.
- Kurtosis
- A measure of how heavy the tails of a distribution are, which shows how often extreme values occur.
- Machine learning
- A branch of AI in which a program learns patterns from data, instead of following rules that a person wrote.
- MapReduce
- A programming model that splits a large data job into map tasks that run in parallel and reduce tasks that combine their results.
- Overfitting
- A model that has learned the noise in its training data, so it scores well there but poorly on new data.
- Pandas
- A Python library for loading, cleaning and analyzing tables of data.
- PCA
- Principal component analysis: a dimensionality reduction technique that combines features into a few new ones that keep most of the variation.
- ROC curve
- A plot of a classifier's true positive rate against its false positive rate at every decision threshold.
- Skewness
- A measure of how lopsided a distribution is, with a longer tail on one side of the mean.
- Underfitting
- A model that is too simple to capture the pattern in the data, so it performs poorly even on its training data.