Gradient Descent Hats
Description: Going downhill fast since 1847 - when Cauchy first described the method every neural network still runs on. A retro academic club badge for the Gradient Descent Club, with the loss curve rolling toward the minimum. For ML engineers and data scientists who get the joke and live the joke.
Description: A stark monochrome composition featuring the text "I HAD POTENTIAL" at the top. A stick figure with arms raised in triumph stands at the apex of a downward sloping curve. Another stick figure sits dejectedly at the bottom, holding a cup. The slope is rendered with a gradient of dots.
Description: The pun "Yes We GAN" alludes to a type of algorithm for unsupervised machine learning, the so-called Generative Adversarial Network (GAN). GANs consist of two artificial neural networks: the generator (G) and the discriminator (D). The generator tries to generate, for example, photorealistic images, while the discriminator tries to identify these "fakes" and tell them apart from real images. Both networks use deep learning to get better at their respective jobs, until the generator is able to create images that look deceptively real. This humorous design is perfect for anyone who is interested in deep learning, artificial intelligence, machine learning, computer science, pattern recognition and technology. A great gift for computer scientis
Description: Show off your geeky side with this "I Love Neural Networks" design featuring a (mostly) fully connected neural network in the shape of a heart! Wear this on your next artificial intelligence conference or gift it to a nerdy programmer friend because deep learning is the future!
Description: The pun "Yes We GAN" alludes to a type of algorithm for unsupervised machine learning, the so-called Generative Adversarial Network (GAN). GANs consist of two artificial neural networks: the generator (G) and the discriminator (D). The generator tries to generate, for example, photorealistic images, while the discriminator tries to identify these "fakes" and tell them apart from real images. Both networks use deep learning to get better at their respective jobs, until the generator is able to create images that look deceptively real. This humorous design is perfect for anyone who is interested in deep learning, artificial intelligence, machine learning, computer science, pattern recognition and technology. A great gift for computer scientis
