Machine Learning illustration pack (40 illustrations)

Machine Learning pack cover

♛ PRO 40 icons in this pack · included with Pro

  • Every Pro pack, plus future ones
  • Icon editor and recolour
  • Molecules, Proteins, Gel Builder and Well Plate exports
  • Use in papers, talks and commercial work

$49 one-time · lifetime (₹2,999 in India) · Get Pro

Pro keeps the free library free.

Machine Learning is a Sciicons Pro illustration pack of 40 original, scientifically accurate illustrations. Machine learning and AI for papers, posters, slides and teaching: the perceptron, fully connected networks, activation functions (ReLU, sigmoid, tanh), backpropagation, loss curves, gradient descent and dropout; architectures (CNN, convolution kernel, LSTM cell, transformer block, self-attention heads, encoder–decoder, autoencoder, GAN, diffusion model); classic ML (decision tree, random forest, SVM, k-means, k-NN, linear and logistic regression, PCA); training and evaluation (train/validation/test split, k-fold cross-validation, overfitting and underfitting, confusion matrix, ROC and precision–recall curves, hyperparameter tuning); data and representation (dataset table, word embeddings, t-SNE/UMAP, tokenization, data augmentation); and AI for science (protein structure prediction, molecular graph neural network, cell segmentation, LLM chat assistant, GPU cluster).

Flat, minimal diagrams on a 512 × 512 canvas with one shared palette: navy and soft blue, warm orange, yellow and green. Every part sits in a named group so Recolor and Edit can target it.

Topics: machine learning, deep learning, artificial intelligence, ai, neural network, perceptron, activation function, relu, sigmoid, backpropagation, gradient descent, loss curve, dropout, cnn, convolution, lstm, rnn, transformer, attention, encoder decoder, autoencoder, gan, diffusion model, decision tree, random forest, svm, k-means, knn, regression, pca, cross validation, overfitting, confusion matrix, roc curve, precision recall, hyperparameter tuning, dataset, embeddings, t-sne, umap, tokenization, data augmentation, alphafold, protein structure prediction, graph neural network, cell segmentation, llm, chatbot, gpu cluster.

Neural network basics (7)

Architectures (9)

Classic machine learning (7)

Training & evaluation (7)

Data & representation (5)

AI for science (5)

License

Included with Sciicons Pro (one-time lifetime purchase). Use the illustrations in papers, theses, posters, slides, grant figures and teaching; editable SVG with named layers. Redistribution of the files as an icon set is not allowed. See how to cite Sciicons.

Related packs

All illustration packs · Browse by category