Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteA bathroom faucet offers a useful first picture of supervised neural-network training: you choose a target water temperature, observe the actual output, measure the difference, and adjust the controls before trying again. The comparison captures the feedback loop, but a neural network uses calculated losses and gradients—not a hand turning a handle.
The faucet-to-neural-network mapping
Imagine a shower with separate hot and cold handles. You want water at a particular temperature. You open both handles, feel the result, decide whether it is too hot or too cold, change the settings, and check again.
Bill Schmarzo uses this scenario to introduce backpropagation and stochastic gradient descent in his 2019 explanation, “Using a Bathroom Faucet to Teach Neural Network Basic Concepts.” The mapping is:
| Faucet experience | Neural-network concept |
|---|---|
| Desired water temperature | Target (the expected output for a training example) |
| Water flowing from the tap | Model prediction produced by a forward pass |
| Difference between desired and actual temperature | Prediction error or loss |
| Changing the hot and cold handles | Updating learned parameters |
| Repeated checks and adjustments | Iterative optimization during training |
Schmarzo summarizes the intended correspondence this way: “The goal of the faucet Neural Network is to find my optimal water temperature by tuning the faucet (model) hyperparameters (weights and biases).” The phrase is memorable, although in technical usage weights and biases are learned parameters, while hyperparameters usually include settings such as the learning rate.
#1 Best Overall
- CRISP CLARITY: This 23.8″ Philips V line monitor delivers crisp Full HD 1920x1080 visuals. Enjoy movies, shows and videos with remarkable detail
- INCREDIBLE CONTRAST: The VA panel produces brighter whites and deeper blacks. You get true-to-life images and more gradients with 16.7 million colors
- THE PERFECT VIEW: The 178/178 degree extra wide viewing angle prevents the shifting of colors when viewed from an offset angle, so you always get consistent colors
- WORK SEAMLESSLY: This sleek monitor is virtually bezel-free on three sides, so the screen looks even bigger for the viewer. This minimalistic design also allows for seamless multi-monitor setups that enhance your workflow and boost productivity
- A BETTER READING EXPERIENCE: For busy office workers, EasyRead mode provides a more paper-like experience for when viewing lengthy documents
How the training loop works
1. Set a target
You begin with a desired temperature. In supervised learning, each training example includes a target output. For an image classifier, for example, the target might be “cat”; for a numerical prediction, it might be a measured value. The target is what the model is trying to match, not an instruction that changes after every prediction. Carnegie Mellon’s curricular material describes training as comparing outputs with expected results: Curricular Modules.
2. Produce an output
Opening the faucet produces water at an actual temperature. A neural network performs a forward pass (also called feed-forward computation): information moves from the input layer through connected calculations to a prediction. NVIDIA distinguishes this learned-parameter stage from later use of the model for inference in its overview of an artificial neural network.
3. Measure the mismatch
If the water is 10 degrees colder than your target, the outcome is wrong by that amount in this simple illustration. A training system computes a loss function that measures the mismatch between prediction and target. “Too hot” or “too cold” provides intuitive direction, but a real loss is a defined mathematical objective that can have a different scale and shape depending on the task.
Rank #2
- Clear visuals. Fluid motion: A 144Hz refresh rate and 1ms MPRT deliver smooth, tear‑free motion across work, gaming, and streaming for clearer, more fluid viewing.
- Eye comfort: TÜV Rheinland 3‑star* certification reduces harmful blue light while preserving stunning color quality without compromise. *TÜV Rheinland 3-star eye comfort certification.
- Wide viewing angle: Get consistent views across a wide 178° /178° viewing angle.
- In-Plane Switching (IPS): See excellent color accuracy and consistency across wide viewing angles with In-plane Switching (IPS) technology.
- Ultra-thin bezels: Maximize your viewing experience with thin bezels.
4. Calculate how parameters contributed
With a faucet, you use experience to decide which handle to move. In a network, backpropagation applies the chain rule to propagate derivative information backward from the loss through the layers. It estimates how changing each parameter would affect the loss. Carnegie Mellon explains feed-forward and backpropagation as separate directions of computation in its educational modules.
5. Update the parameters
An optimizer such as gradient descent uses the calculated gradients to choose parameter changes intended to reduce loss. Stochastic gradient descent performs such updates using one example or a small batch at a time, rather than waiting to process the entire training set. Backpropagation calculates gradient information; gradient descent (or another optimizer) uses it. They are related, but not synonyms.
6. Control the step size
The learning rate determines how large an update is. A larger rate can move faster, but it can overshoot or fail to converge correctly; a very small rate may require many updates. Carnegie Mellon discusses this trade-off in its training material. Turning a faucet handle a little versus a lot is an intuition for step size, not a literal implementation of a learning-rate schedule.
Rank #3
- ALL-EXPANSIVE VIEW: The three-sided borderless display brings a clean and modern aesthetic to any working environment; In a multi-monitor setup, the displays line up seamlessly for a virtually gapless view without distractions
- SYNCHRONIZED ACTION: AMD FreeSync keeps your monitor and graphics card refresh rate in sync to reduce image tearing; Watch movies and play games without any interruptions; Even fast scenes look seamless and smooth.
- SEAMLESS, SMOOTH VISUALS: The 75Hz refresh rate ensures every frame on screen moves smoothly for fluid scenes without lag; Whether finalizing a work presentation, watching a video or playing a game, content is projected without any ghosting effect
- MORE GAMING POWER: Optimized game settings instantly give you the edge; View games with vivid color and greater image contrast to spot enemies hiding in the dark; Game Mode adjusts any game to fill your screen with every detail in view
- SUPERIOR EYE CARE: Advanced eye comfort technology reduces eye strain for less strenuous extended computing; Flicker Free technology continuously removes tiring and irritating screen flicker, while Eye Saver Mode minimizes emitted blue light
What a single neuron is doing
The faucet story describes a whole feedback loop. A basic neuron explains the calculation inside that loop:
- It receives input values.
- It multiplies each input by a weight.
- It adds the weighted values and a bias.
- It applies an activation function to the result.
In compact form, a neuron first computes a weighted sum such as z = w₁x₁ + w₂x₂ + … + b, then produces an activation from z. A weight controls how strongly an input (or a preceding neuron’s output) influences the calculation. A bias supplies an adjustable offset. An activation function transforms the weighted input and, across layers, enables nonlinear behavior. Microsoft’s archived explanation, “Test Run – Dive into Neural Networks,” defines these building blocks; IBM provides a modern overview in “What Is a Neural Network?”
What the faucet analogy captures—and what it leaves out
| The analogy captures | It omits or simplifies |
|---|---|
| A target output | The choice and mathematics of a loss function |
| An observed prediction | Many inputs, layers, neurons, and coupled parameters |
| Error-based feedback | Derivative calculations through every operation |
| Repeated adjustments | Data batching, regularization, validation, and other training details |
| Eventually using tuned settings | The engineering steps needed to deploy and monitor a model |
Do not assign one faucet handle to one specific neural-network weight. A shower has a few controls and one immediately observed scalar outcome; a practical network may contain many interconnected parameters whose effects interact across layers.
Rank #4
- CRISP CLARITY: This 22 inch class (21.5″ viewable) Philips V line monitor delivers crisp Full HD 1920x1080 visuals. Enjoy movies, shows and videos with remarkable detail
- 100HZ FAST REFRESH RATE: 100Hz brings your favorite movies and video games to life. Stream, binge, and play effortlessly
- SMOOTH ACTION WITH ADAPTIVE-SYNC: Adaptive-Sync technology ensures fluid action sequences and rapid response time. Every frame will be rendered smoothly with crystal clarity and without stutter
- INCREDIBLE CONTRAST: The VA panel produces brighter whites and deeper blacks. You get true-to-life images and more gradients with 16.7 million colors
- THE PERFECT VIEW: The 178/178 degree extra wide viewing angle prevents the shifting of colors when viewed from an offset angle, so you always get consistent colors
Likewise, your hand feeling the water is not backpropagation. Human feedback reports an outcome to a person. Backpropagation is an algorithmic calculation of derivatives. A model also does not update itself merely because it “sees” a result: training requires examples, targets, a specified loss, and an optimization procedure.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Training versus inference
During training, the system has target outputs and changes its parameters to reduce loss. After training, those learned parameters are normally held fixed while the model processes new, unseen inputs. That use stage is inference. Carnegie Mellon describes the distinction between tuning parameters from examples and applying the resulting network to new data, and NVIDIA makes the same training-versus-inference distinction in its neural-network overview: Artificial Neural Network.
In faucet terms, training is the repeated handle adjustment needed to reach a useful setting. Inference is opening the faucet at the learned setting and using the resulting output; there is no target supplied to update the model during that ordinary prediction.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Best Value
- Incredible Images: The Acer KB272 G0bi 27" monitor with 1920 x 1080 Full HD resolution in a 16:9 aspect ratio presents stunning, high-quality images with excellent detail.
- Adaptive-Sync Support: Get fast refresh rates thanks to the Adaptive-Sync Support (FreeSync Compatible) product that matches the refresh rate of your monitor with your graphics card. The result is a smooth, tear-free experience in gaming and video playback applications.
- Responsive!!: Fast response time of 1ms enhances the experience. No matter the fast-moving action or any dramatic transitions will be all rendered smoothly without the annoying effects of smearing or ghosting. A 120Hz refresh rate speeds up the frames per second to deliver smooth 2D motion scenes in gaming and video.
- 27" Full HD (1920 x 1080) Widescreen IPS Monitor | Adaptive-Sync Support (FreeSync Compatible)
- Refresh Rate: Up to 120Hz | Response Time: 1ms VRB | Brightness: 250 nits | Pixel Pitch: 0.311mm
A compact worked intuition
Suppose the target is 38 °C and the first mixture is 32 °C. The six-degree mismatch is evidence that the current settings are not adequate. You adjust the controls and obtain 36 °C, then 37.5 °C. This sequence illustrates decreasing error through iterative updates.
A neural network follows the same high-level pattern: inputs produce a prediction, a loss compares it with the target, backpropagation computes parameter sensitivities, and an optimizer applies an update. The numerical update is not based on a person’s qualitative sensation; it is based on gradients and the chosen learning rate. Nor does iterative descent guarantee a global optimum; it is a method for seeking lower loss.
Key terms in plain language
- Input: Information supplied to the model.
- Weight: A learned number that controls the influence of an input or earlier neuron output.
- Bias: A learned offset added to a weighted sum.
- Weighted sum: Inputs multiplied by weights, combined, and shifted by a bias.
- Activation function: A transformation applied after the weighted sum that helps a network represent nonlinear relationships.
- Forward pass (feed-forward): Computation from inputs toward a prediction.
- Loss: A numerical measure of disagreement between prediction and target.
- Backpropagation: Backward propagation of derivative information used to determine how parameters affect loss.
- Gradient descent: An optimization method that uses gradients to select parameter updates.
- Learning rate: The setting that controls update size.
- Inference: Applying learned parameters to produce predictions on new inputs.
How to use the metaphor responsibly
- Use it to introduce targets, predictions, error, and iterative adjustment.
- Explain that the faucet gives directional intuition, while the network computes a formal loss.
- Keep backpropagation (gradient calculation) separate from the optimizer (parameter update).
- State that real networks have many interacting parameters and layers.
- Present the faucet as an explanatory device, not evidence that this teaching method has been empirically validated.
The faucet is therefore a good doorway into neural-network training: it makes the feedback loop tangible without pretending that a neural network learns by consciously turning two handles.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
