AI and Data Bias Quiz

A company developed a smartphone app that used GPS and accelerometer data to identify and report potholes in a city’s streets. Their intent was to help allocate resources to patch thousands of potholes. However, many of the city’s elderly and lower-income residents did not have smartphones, resulting in datasets biased in favor of more affluent neighborhoods.

Which type of bias is this an example of?

• A. Human cognitive bias

• B. Data collection bias

• C. Automation bias

• D. Reinforcement bias

Answer explanation

Bias ini terjadi karena metode pengumpulan data hanya mengandalkan smartphone. Akibatnya, data yang dikumpulkan tidak mewakili seluruh populasi (terutama kelompok lansia dan berpenghasilan rendah), sehingga terjadi bias dalam dataset.

2. A bank builds an AI that predicts whether a customer is likely to default on a loan. The bank intends to use the AI predictions to approve or deny loan applications. Some customers may dispute the loan decisions. The bank wants to make sure that it keeps appropriate records to explain to customers how decisions are made.

Which information should the bank use to mitigate future disputes?

• A. The engineer who wrote the prediction code.

• B. The number of training experiments that were run to select the best algorithm.

• C. The dataset that was used to train the AI.

• D. The number of predictions the AI has made in production.

Answer explanation

Dataset yang digunakan untuk melatih AI adalah informasi penting untuk menjelaskan kepada pelanggan bagaimana keputusan dibuat. Dengan mengetahui dataset tersebut, bank dapat mengaudit, memeriksa, dan membuktikan bahwa keputusan AI didasarkan pada data yang relevan dan adil.

You are building an AI to predict whether a loan applicant should be approved or denied for a loan. The data has many features because your company knows a lot about its customers. You want to start with only the most-valuable features to give the AI its best chance for success.

How should you do this?

• A. Talk with the people who manually approve the loans and ask them what factors to consider.

• B. Ask the customer what they think would be the best things about them to consider when evaluating the application.

• C. Keep only the features that are numbers (like the customer's age) because AIs only work with numbers.

• D. Keep only the features that describe personal things about the customer because those are the most distinctive.

Answer explanation

Bias ini terjadi karena metode pengumpulan data hanya mengandalkan smartphone. Akibatnya, data yang dikumpulkan tidak mewakili seluruh populasi (terutama kelompok lansia dan berpenghasilan rendah), sehingga terjadi bias dalam dataset.

What is a vulnerability that a malicious actor can exploit to attack or undermine an AI system?

• A. Overfitting.

• B. Unsupervised learning.

• C. Data encryption.

• D. Data extraction.

Which two factors are likely to be important elements in making a choice of dataset to meet this goal?(Choose 2)

• A. Missing elements or the presence of corrupt elements.

• B. Whether the dataset was located by using a dataset search engine.

• C. Ensuring that the dataset is sufficient for the purpose.

• D. Whether the dataset was released by an academic or commercial entity.

• E. The file format of dataset elements such as images.

Answer explanation

Dalam memilih dataset untuk mencapai tujuan model:

Penting untuk memastikan tidak ada data yang hilang atau rusak (A), karena ini memengaruhi kualitas pelatihan.

you are trying to create a model to estimate the land area covered by ice in the Artic circle. you have 100.000 labeled satellite image of the artic circle. You can use the images to train a machine-learning model to distinguish between ice, water, and land in order to create your land area estimates.

how should you allocate images to the training and testing dataset?

• A. Use all the data to train the model and use a subset of the data for testing

• B. Select approximately 80% of the images for training and use the rest for testing

• C. Train and test the model on 100% of the labeled images

• D. Select approximately 50% of the images for training and use the rest for testing

Answer explanation

Dalam machine learning, praktik terbaik adalah membagi dataset menjadi training dan testing set, biasanya 80/20 atau 70/30.

Menggunakan 100% data untuk training (opsi C) tidak memungkinkan evaluasi model yang obyektif.

The question asks why subject-matter experts without technical expertise should be included in an AI project team.

• A. To evaluate tools used to integrate AI into existing applications.

• B. To train them in applied AI and data engineering skills.

• C. To organize and cleanse the structured and unstructured data in a custom AI solution.

• D. To feed domain-specific data into the algorithm so the AI system is efficient and unbiased.

Answer explanation

Subject-matter experts yang bukan teknisi dibutuhkan dalam tim AI karena:

Mereka memiliki pengetahuan mendalam tentang domain (misalnya perbankan, kesehatan).

Mereka memastikan data dan keputusan yang digunakan oleh AI relevan, akurat, dan tidak bias.

Tanpa mereka, AI dapat salah memahami konteks atau menghasilkan hasil yang tidak sesuai kebutuhan bisnis.

The question asks what important factor, besides availability of feedback and volume of data, should be used when determining how often a model should be retrained, considering periodic, performance-based, or data-change-based retraining strategies. The options are:

• A. Resource cost

• B. Performance of the model in production

• C. Ease of implementation

• D. Periodicity

Answer explanation

Pertanyaan ini membahas frekuensi retraining model agar tetap relevan dan berkinerja optimal. Selain mempertimbangkan ketersediaan feedback dan volume data, Anda juga perlu melihat periodicity (keteraturan atau frekuensi pembaruan data dan perubahan lingkungan) untuk menentukan kapan model harus dilatih ulang.

Which three benefits should you expect to gain by receiving this user feedback?(Choose 3)

• A. Insights on user satisfaction with the system.

• B. Opportunities to profile users more easily.

• C. Effective A/B testing of different versions of the recommendation algorithm.

• D. Insights on the latest technology developments.

• E. Reduced model size.

Answer explanation

B: Feedback langsung (thumbs up/down) memberikan wawasan tentang seberapa puas pengguna dengan rekomendasi yang diberikan.

C: Feedback membantu memprofilkan preferensi pengguna sehingga AI dapat memberikan rekomendasi yang lebih personal.

D: Feedback juga memungkinkan A/B testing untuk membandingkan berbagai versi algoritma rekomendasi dan memilih yang terbaik.

Pilihan A (insight teknologi terbaru) dan E (model size lebih kecil) tidak berhubungan langsung dengan manfaat dari user feedback.

What should your adoption plan contain?

• A. Cloud Storage= Costs vary by the amount of space needed at a given time, There are no extra costs for maintenance and upgrades.

  Local Server= Involves the cost of installing and maintaining hardware, It is accessible even without an internet connection.

• B. Cloud Storage= Involves the cost of installing and maintaining hardware, It is accessible even without an internet connection.

  Local Server= Costs vary by the amount of space needed at a given time, There are no extra costs for maintenance and upgrades.

• C. Cloud Storage= Costs vary by the amount of space needed at a given time, Involves the cost of installing and maintaining hardware,

  Local Server= It is accessible even without an internet connection. , There are no extra costs for maintenance and upgrades.

• D. Cloud Storage= Involves the cost of installing and maintaining hardware, There are no extra costs for maintenance and upgrades.

  Local Server= Costs vary by the amount of space needed at a given time, It is accessible even without an internet connection.

Answer explanation

Cloud Storage: Biaya fleksibel sesuai kapasitas yang digunakan dan tidak perlu biaya tambahan untuk pemeliharaan karena dikelola oleh penyedia cloud.

Local Server: Memerlukan biaya instalasi dan pemeliharaan perangkat keras tetapi dapat diakses tanpa koneksi internet.